JPH04321214A - Position detector - Google Patents

Abstract

PURPOSE:To obtain a position detector which can accurately detect a position by using a plurality of positional information obtained by luminous fluxes having different spectral characteristics of detecting positions of alignment marks and imaging means on a mask. CONSTITUTION:When imaging means 14 and alignment marks 5 of a second article are position-detected through a projection lens 1 for projecting a pattern on a first article on the second article, alignment images based on a plurality of luminous fluxes having different spectral distributions are sequentially formed on the means 14 by using an illumination system having a wavelength selecting means 12 having a variable wavelength range. Positional information of the alignment mark images based on the plurality of the fluxes formed on the means 14 are respectively stored in memory means, and the plurality of the positional information stored in the memory means are utilized.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a position detection device, and in particular, when an electronic circuit pattern on a reticle surface (first object surface) is projected and transferred onto a wafer surface (second object surface) using a projection lens. The present invention relates to a position detection device suitable for an exposure apparatus for manufacturing semiconductor devices, which detects the relative position of an imaging means that is in an optically conjugate relationship with the wafer and is aligned with the reticle in advance. be.

0002

2. Description of the Related Art Conventionally, in exposure apparatuses for manufacturing semiconductor devices, one method for detecting the relative position (alignment) between a reticle and a wafer is a TTL method using a projection lens. This TTL method further includes the so-called on-
There are two methods: the Axis method and the off-Axis method. In most cases, these two types of position detection devices use monochromatic light or a light beam whose full width at half maximum is narrowed to several nanometers as illumination light for observing alignment marks during position detection (alignment light).
It is used as

As described above, since the wavelength width of the light beam used in conventional position detection devices is narrow, for example, when observing an alignment mark on a wafer surface coated with resist, interference fringes are generated due to light reflected from the resist surface and the substrate surface. occurs a lot,
This causes detection errors.

In order to reduce interference fringes at this time, an exposure apparatus has been proposed in which the alignment mark is observed using a light source that emits a polychromatic light beam with a wide spectrum width and a full width at half maximum of approximately several tens of nanometers. .

[0005]

Generally, when a projection optical system is constructed from a refractive optical system, chromatic aberration occurs due to chromatic dispersion of the material, which is one of the causes of lowering position detection accuracy. In many cases, various aberrations of the projection optical system are optimized by the exposure wavelength. For this reason, if the exposure wavelength and the alignment wavelength are different, a lot of chromatic aberration will occur at the alignment wavelength when the projection lens is used. Furthermore, when polychromatic light (wavelengths λ0 to λn) consisting of a wavelength range different from the exposure wavelength is used, the amount of various aberrations generated when passing through the projection lens is different for each wavelength from wavelength λ0 to wavelength λn. It's coming.

[0006] For this reason, the image quality of an image obtained by combining images based on the light beams of each of these wavelengths, that is, an alignment mark image of polychromatic light, is lower than that of a monochromatic light image, although there are no interference fringes. (In addition, the present applicant has proposed in Japanese Patent Application No. 1-198261 an exposure device that effectively corrects aberrations when observing an alignment mark image using a TTL method using a polychromatic light source.) When a light beam with a wide wavelength width is used, the difference between an alignment wavelength for which aberrations are well corrected by a positioning microscope or the like and an alignment wavelength for which the aberrations are not well corrected increases.

[0007] When using a refractive optical system as described above, in order to reduce the influence of interference fringes, it is sufficient to widen the wavelength width.
On the other hand, if the lens is made too wide, the amount of various aberrations will increase and the image quality will deteriorate, so it cannot be made too wide.

In addition, when observing alignment marks on the wafer surface using polychromatic light, even if the wafer surface is irradiated with illumination light with uniform intensity of each wavelength, interference due to resist film thickness, resist surface and The spectral intensity distribution of an alignment mark image obtained by an imaging means such as a CCD changes due to the influence of the reflectance of the substrate surface and the like. That is, the amount of aberration generated by the optical system when reproducing a wafer image varies depending on the conditions on the wafer surface (differences in reflectance, shape, etc.).

If the position is detected using the TTL method using polychromatic light in this way, the occurrence of interference fringes will be reduced when monochromatic light is used, but the alignment mark image will be distorted due to the occurrence of various aberrations due to the widening of the wavelength width. There was a problem in that the observation condition deteriorated and sufficient position detection accuracy could not be obtained.

Further, there are problems in that the detection accuracy varies from shot to shot or from wafer to wafer, and the offset is not constant, making it difficult to obtain high detection accuracy.

The present invention uses polychromatic light with a wide wavelength range to detect a relative position between a reference position on a predetermined surface and an alignment mark on a wafer surface. By using the position information obtained from the
The purpose of the present invention is to provide a position detection device with high detection accuracy.

0012

[Means for Solving the Problems] The position detection device of the present invention detects a reference position on a predetermined surface and the second object surface through a projection lens that projects a pattern on a first object surface onto a second object surface. When detecting the position relative to the alignment mark formed above, an illumination system having a wavelength selection member with a variable wavelength range is used to generate an alignment mark image based on a plurality of light beams having mutually different spectral distributions. sequentially formed on a predetermined surface, and storing positional information for each alignment mark image based on the plurality of light beams formed on the predetermined surface in a memory means;
It is characterized in that position detection is performed using the plurality of position information stored in the memory means.

[0013]

Embodiment FIG. 1 is a schematic diagram of a main part of Embodiment 1 of the present invention. First, the illumination of the alignment mark 5 on the wafer 4 surface or the fiducial mark 24 on the θz stage and their alignment will be explained.

1 is a projection lens, and an illumination system (not shown)
The electronic circuit pattern on the six surfaces of the reticle, illuminated with exposure light, is projected onto the four surfaces of the wafer. 3 is a θz stage on which a wafer 4 is placed. 2 is an XY stage on which the θz stage 3 is mounted. θz stage 3
The wafer 4 is controlled to be driven in any direction by the XY stage 2 and the XY stage 2. 5 is an alignment mark, and wafer 4
It is formed on the surface. Reference numeral 16 denotes a light source means, which emits polychromatic light with a wide wavelength range. Reference numeral 15a denotes a bandpass filter, which has a spectral characteristic of transmitting wavelengths λ0 to λn as shown in FIG. 2, for example. The center wavelength λC of the transmission wavelength range of the bandpass filter 15a is the wafer 4
It has a non-exposure wavelength that is not sensitive to the resist applied on the surface, for example 63, which is the same as the oscillation wavelength of a He-Ne laser.
It is 3 nm.

Reference numeral 15 denotes an illumination lens, which condenses the polychromatic light beam that has passed through the bandpass filter 15a. 12
is a wavelength selection member, for example, as shown in FIG.
2b is provided with a plurality of color filters 12a1 to 12a8 in a turret type manner. Color filters 12a1 to 12a
8 passes a light beam having specific spectral characteristics f1 to f8 with a narrow wavelength width as shown in FIG. 4, for example. 13
is a driving means, which rotates the wavelength selection member 12 so that a predetermined color filter is positioned on the optical path.

Reference numeral 10 denotes a beam splitter, which reflects a beam of light having a specific spectral characteristic that has passed through the color filter of the wavelength selection member 12 toward the objective lens 9. A correction optical system 8 corrects various aberrations generated in the projection lens 1, such as coma and astigmatism. Reference numeral 7 denotes a mirror, and its angle is set so that it is telecentric at the observation image height of the alignment mark 5 on the wafer 4 surface. Reference numeral 11 is a relay lens, and reference numeral 14 is an imaging means, which is comprised of, for example, a CCD camera.

21 is a bar mirror for an interference distance meter;
It is placed on two sides of the XY stage. Reference numeral 22 denotes an interference distance meter, which measures the amount of movement of the XY stage 2. 23
is a control device, which drives and controls the XY stage 2 based on a signal from the interference distance meter 22, for example.

24 is a fiducial mark, θ
They are provided on the surface of the z stage 3, and in this embodiment, as shown in FIG. 8, they are provided in a plurality of regions 81, 82, and 83 so that they can be observed with each optical system described later. One fiducial mark 24a is made up of four types of marks 51a, 51b, 51c, and 51d as shown in FIG. 5, for example. The fiducial mark in FIG. 5 is drawn using a reflective material such as chrome, and in the figure, for example, the white area is the reflective area.

In this embodiment, the polychromatic light from the light source means 16 is filtered by a bandpass filter 15a to have wavelengths λ0 to λ0 shown in FIG.
A light beam having a wavelength λn in a relatively wide wavelength range is passed through. Then, the light is focused by the illumination lens 15 and the wavelength selection member 12
One color filter 12ai allows a light beam having a narrow wavelength range spectral characteristic (fi) shown in FIG. 4 to pass through. Then, after being reflected by the beam splitter 10, the objective lens 9, the correction optical system 8, the mirror 7, and the projection lens 1
The fiducial mark 24 and the alignment mark 5 on the surface of the wafer 4 are irradiated through the beam. For example, the reflected light from the alignment mark 5 on the wafer 4 surface returns to the original optical path, that is, it passes through the projection lens 1, the mirror 7, the correction optical system 8, the objective lens 9, and the beam splitter 10 in order, and is transmitted through the relay lens 11. The light enters the imaging means 14 and forms an alignment mark image on its surface.

In this embodiment, the axial chromatic aberration generated by the optical system consisting of each element from the projection lens 1 to the relay lens 11 is a constant aberration centered around the center wavelength λC of the bandpass filter 15a. It has been corrected so that

In this embodiment, an illumination system is composed of an optical system extending from the light source means 16 to the wafer 4 via the projection lens 1 and a system extending from the wafer 4 to the imaging means 14 via the projection lens 1.

Next, the illumination observation and alignment of the reticle reference marks 32, 32a placed near the reticle 6 and the wafer surface 4 will be explained.

Reference numerals 30 and 30a are fibers, which emit a light beam having approximately the same wavelength as the exposure light from an illumination system (not shown), and a reticle reference mark 32 placed near the reticle 6 via prism reflectors 31 and 31a. , 32a. Reference numerals 33 and 33a are mirrors that reflect the light beams from the reticle reference marks 32 and 32a to the objective lenses 34 and 33a.
34a. The mirror 33 (33a) and the objective lens 34 (34a) can move together on a plane parallel to the reticle 6.

35 and 35a are beam splitters. Reference numerals 36 and 36a are relay lenses, which use the light flux that has passed through the beam splitters 35 and 35a to place a reticle reference mark 3 on the surface of an image pickup device 37 and 37a of a CCD camera, etc.
2,32a is formed.

Reference numerals 39 and 39a denote fibers, which emit a light beam having a wavelength substantially equal to the exposure light from an illumination system (not shown), and input the light beam into observation illumination lenses 38 and 38a. In the figure, the distance between the reticle reference marks 32, 32a and the reticle 6 is set to be within the depth of focus of the objective lens 24.

In this embodiment, each element 33 to 37 (33a to
37a) constitutes an optical system 101 (101a), and each element 7 to 14 constitutes an optical system 102.

The reticle reference marks 32, 32a are emitted from the fibers 30, 30a and are reflected by the prism reflectors 31, 31.
Transmissive illumination is performed using a light beam deflected by a. The light beam transmitted through the reticle reference marks 32 and 32a is the reticle 6
, the mirrors 33, 33a, and the objective lens 34 mentioned above.
, 34a, beam splitters 35, 35a, and image pickup elements 37, 37a via relay lenses 36, 36a.
A reticle reference mark image is formed on the surface of the reticle. This allows the reticle reference mark image to be observed. At this time, the reticle 6 and the reticle reference marks 32, 32a
Since the pattern surfaces of the two faces each other, it is possible to observe both at the same time.

The light beams emitted from the fibers 39, 39a pass through illumination lenses 38, 38a, and are deflected by beam splitters 35, 35a toward objective lenses 34, 34a. Thereafter, the light passes through the objective lenses 34, 34a, the mirrors 33, 33a, the reticle 6, and the projection lens 1, and is incident on the surface of the wafer 4.

At this time, the pattern on the surface of the reticle 6 is set so that the light beam passes through the reticle 6. Further, each of the above-mentioned elements is adjusted so that the incident light beam becomes approximately telecentric on the wafer 4 surface.

The light beam reflected on the surface of the wafer 4 returns to its original optical path, passes through beam splitters 35, 35a, passes through relay lenses 36, 36a, and enters image pickup devices 37, 37a. At this time, the imaging elements 37, 37a, the reticle pattern surface, and the wafer 4 surface are optically in a conjugate relationship with each other. Therefore, it is possible to simultaneously observe both the reticle surface and the wafer surface on the imaging elements 37, 37a.

Next, a procedure for measuring the baseline (distance between the alignment position and the exposure position) in this embodiment will be explained. (B) Adjust the reticle reference marks 32, 32a in advance so that the center of the reticle 6 coincides with the optical axis of the projection lens 1 when the alignment between the reticle mark on the reticle 6 surface and the reticle reference marks 32, 32a is completed. and attach it to the main body of the device. (b) Reticle 6 (reticle mark) placed on a reticle stage (not shown) and reticle reference marks 32, 32
a is observed by the optical system 101 (101a), and the amount of relative positional deviation between the reticle mark and the reticle reference mark is detected. (c) After detecting the amount of positional deviation, the reticle stage is driven to align the reticle mark and the reticle reference mark. This allows the center of the reticle to coincide with the optical axis of the projection lens 1. (d) The fiducial mark 24 is moved under the projection lens 1 by the XY stage 2. The fiducial mark 24 is attached to each optical system 101, 101a,
102, and the corresponding marks are set to be observed at the same time. (e) The optical systems 101 and 101a detect the amount of relative positional deviation between the reticle mark and each fiducial mark 24, and the XY stage 2 is driven based on the detection result to position both marks. Perform alignment. The amount of movement of the XY stage 2 is measured by an interference distance meter 22. (f) At this time, the position of the fiducial mark image projected onto the surface of the image sensor 14 of the optical system 102 is measured. This allows baseline measurements to be made.

During the subsequent printing sequence, the alignment mark 5 on the surface of the wafer 4 is placed at a predetermined position on the surface of the imaging means 14, thereby performing alignment.

Next, the position detection method (alignment method) of this embodiment will be explained.

In this embodiment, before measuring the alignment marks on the four wafer surfaces and exposing the wafer surface, the distance (baseline) between the alignment position and the exposure position is periodically measured.
We are measuring and correcting changes over time. The fiducial mark 24 shown in FIG. 5 described above is used as the measurement mark. The focus of the fiducial mark 24 is measured and corrected by a sensor (not shown). The projected waveforms of the fiducial marks 24 formed on the surface of the imaging means 14 are, for example, marks 51a to 51, respectively, as shown in FIG.
Waveforms 61a, 61b, 61c, 61d corresponding to d, respectively.
It looks like this.

In this embodiment, four types of marks 51a to 51 are used.
d, the color filter 1 of the wavelength selection member 12 of the illumination system
The light is sequentially illuminated through 2a1 to 12a8, and the position on the surface of the imaging means 14 is measured. At this time, the XY stage 2 is monitored by the interference distance meter 22 so that it remains in a stopped state. The measured values of the marks 51a to 51d through the color filters 12a1 to 12a8, that is, the measured values based on specific spectral characteristics, are stored in a memory means in the control device 23, for example. The measured value at this time is treated as not changing until the next baseline measurement. At the same time, for example, when the mark 51a in FIG. 5 is measured, the brightness in the area near the mark used for mark measurement, such as the output from the imaging means 14, is also stored in the memory means for each color filter. The position M of the fiducial mark 24 is calculated from the following equation based on the measured value from the memory means.

[0036]

[Equation 1] However, Ii is the amount of light reflected from the fiducial mark when the fiducial mark 24 is illuminated with the color filter 12ai, and mi is the surface of the imaging means 14 of the fiducial mark when it is illuminated with the color filter 12ai. This is the position measurement value at the top. The position M obtained by equation (1) is used for baseline correction. For example, the amount of reflected light I from the fiducial mark 24 is
In the case of values of f1 to f8 as shown in FIG. 7 for every 2a8, the position M is determined by applying equation (1).

Next, the alignment mark 5 on the 4th surface of the wafer
The position on the imaging means 14 surface is detected. Regarding the alignment mark 5 on the surface of the wafer 4, the position measurement value Wi and brightness Li on the surface of the imaging means 14 in the light flux that has passed through each color filter 12a1 to 12a8 of the wavelength selection member 12 are determined.

When detecting the position of the alignment mark 5, in addition to the aberration of the optical system, the level difference, reflectance, and resist of the wafer itself affect the measured value.

In this embodiment, in order to remove the influence of aberrations from the alignment mark measurement values, the projected waveform for each color filter is calculated from among the measurement values of the fiducial mark 24 that have been previously obtained and stored in the memory means. The next process is performed using the measurement value of the most similar mark as a numerical value for correction.

[0040]

##EQU00002## where Li is the amount of light reflected from the wafer alignment mark when the alignment mark on the wafer surface is illuminated by the color filter 12ai, and ma is the average value of the measured values mi at the fiducial mark.

In this embodiment, the measurement error caused by the aberration of the projection lens during position detection on the imaging means surface when the alignment mark on the wafer surface is illuminated with polychromatic light through the projection lens is eliminated using the method described above. I'm doing less.

If such measurement is performed, for example, on the first wafer in a lot, the substantial decrease in throughput can be almost ignored. Incidentally, in this embodiment, even if the wavelength selection member 12 is disposed between the relay lens 11 and the imaging means 14, the same effect as described above can be obtained.

In the present invention, when the wavelength for illuminating the alignment mark becomes distant from the wavelength at which the aberrations of the observation system for position detection are corrected, the amount of aberration increases and measurement accuracy deteriorates. Therefore, it is preferable to weight the above-mentioned equation (2) so that the contribution of the color filter in the wavelength range where the measurement accuracy decreases is reduced. For example, when ai is the measurement accuracy of the alignment mark during illumination when using each color filter 12ai

[0044]

[Math 3] It is better to

Furthermore, in the present invention, one or two disk-shaped filters whose spectral transmittance changes continuously may be used as the wavelength selection member. According to this, the wavelength selection member can be configured in a small size.

Further, in the present invention, a dye laser, for example, which can continuously change and select the wavelength may be used as the light source means. In this case, the dye laser corresponds to the wavelength selection means.

[0047]

According to the present invention, when performing relative position detection between a reference position on a predetermined surface and an alignment mark on a wafer surface using polychromatic light having a wide wavelength width, the spectral characteristics By using position information obtained from a plurality of light beams with different distributions, it is possible to reduce the influence of chromatic aberration of the projection lens and achieve a position detection device with high detection accuracy.

[Brief explanation of drawings]

[Figure 1] Schematic diagram of main parts of Example 1 of the present invention

[Figure 2]
An explanatory diagram of the spectral characteristics of the bandpass filter in Figure 1

[Figure 3] Explanatory diagram of the wavelength selection member in Figure 1

[Figure 4]
An explanatory diagram of the spectral characteristics of the color filter in Figure 3

[Figure 5] Mark explanatory diagram of the fiducial mark in Figure 1

[Figure 6] Output signal waveform diagram from the mark in Figure 5

[Figure 7
] An explanatory diagram of the amount of reflected light when the mark is illuminated with each color filter in Figure 3

[Figure 8] Schematic diagram of the fiducial mark in Figure 1

[Explanation of symbols]

1 Projection lens 2 XY stage 3 θz stage 4 Wafer 5 Alignment mark 6 Reticle 7 Mirror 8 Correction optical system 9 Objective lens 10 Beam splitter 11 Relay lens 12 Wavelength selection member 13 Driving means 14 Imaging means 15 Illumination lens 23 Control device 32 Reticle reference mark 32a Reticle reference mark

Claims (1)

[Claims] Claim 1: A relative position between a reference position on a predetermined surface and an alignment mark formed on the second object surface through a projection lens that projects a pattern on the first object surface onto the second object surface. When performing detection, alignment mark images based on a plurality of light beams having mutually different spectral distributions are sequentially formed on the predetermined surface using an illumination system having a wavelength selection member with a variable wavelength range. The position information for the alignment mark images based on the plurality of light beams formed in the above is stored in a memory means, and the position is detected using the plurality of position information stored in the memory means. Detection device.