JPH03148815A - Focus detection and exposure device using the same - Google Patents

Abstract

PURPOSE:To improve the transfer of fine circuit pattern by a method wherein slits limiting the passage of illumination light are reflected halfway of optical path on a detected surface to detect the position of the slit image for processing the slippage from the focus surface. CONSTITUTION:The signal waveforms of slit images 13a, 13b formed through optical paths 6, 7 reflecting on the surface 1 detected through a linear sensor 18 of an optical system A are taken in a controller 20. Likewise, the other waveforms of the other slit images 13a', 13b' formed through the other optical paths 6', 7' reflecting on the surface 1 detected through the other linear sensor 18' of another optical system B are taken in the controller 20 to be compared with the reference signal waveforms of slit image correct in focus. Finally, the tilt of detected surface 1 is processed from the positional flucturation to drive a tilt stage 27 by this signal while the signal of tilt elements 24-26 are transferred to driving circuits 21-23 to correct the signal in the focus by a reduction projection lens 2 for driving the stage 27 so as to correct the stage 27 for letting the detected surface 1 in the focus.

Description

Detailed Description of the Invention [Industrial Field of Application] The present invention relates to a focus detection method and an exposure apparatus using the detection method, The present invention relates to a focus detection method suitable for detecting displacement at high speed and with high precision, and an exposure apparatus using the detection method.

[Conventional technology]

Conventionally, in a reduction projection exposure apparatus used for manufacturing LSTs, when projecting a circuit pattern on a reticle onto a test surface using a reduction projection lens, the position of the test surface is corrected to adjust the focus allowed by the reduction projection lens. A pattern transfer method is used that prevents blurring caused by defocus by holding the surface to be inspected within the test surface, thereby making it possible to transfer submicron fine patterns.

In addition, when detecting the deviation of the target surface from the focus position of the reduction projection lens using the above pattern transfer method, for example, as described in Japanese Patent Application Laid-Open No. 56-42205,
A light beam with a long and narrow cross section is irradiated obliquely onto the surface to be inspected, and the reflected light from the surface is imaged at the position of a detection slit provided in front of the light-receiving element. By relatively vibrating this formed image and the detection slit, a photoelectric conversion element detects a modulated signal caused by the light that has passed through the detection slit, and by detecting the phase of this modulated signal, A method was used to obtain the signal of the amount of deviation.

According to this focus detection method, a cross-section of a light beam narrowed into a long and narrow rectangular (or linear) beam is irradiated onto the test surface, so when detecting a test surface with a regular pattern engraved on it, the If the length direction of the cross-section of the beam irradiated onto the surface to be inspected coincided with the direction of the pattern, the signal detected from the reflected light was likely to be affected by the pattern, resulting in a decrease in accuracy. For this reason, it is necessary to improve the accuracy of focus detection by averaging the signal by illuminating from a direction different from the regular direction of the pattern and distributing the area illuminated with illumination light across multiple patterns. There were restrictions on the placement of the optical system that performs focus detection.

In addition, in the focus detection method described above, which detects the cross-sectional image of the beam irradiating the test surface, even if the test surface is tilted, the angle of the light beam reflected from the test surface only changes, and the test surface is not irradiated. Since the position of the light beam does not change, the image formed on the detection slit of the cross section of the light beam irradiated on the test surface does not move even if the test surface is tilted, so it cannot be detected as a signal. There have been problems such as the inability to detect a shift from the focal point of the Ii/T projection lens that occurs in the periphery of the test surface due to the test surface being tilted.

[Problem to be solved by the invention]

In the above-mentioned conventional technology, as mentioned above, the cross section of the illumination light beam has a long rectangular (or linear) shape, so the cross section of the portion of the light beam irradiated onto the test surface becomes elongated and has a directional pattern. In the case of a surface to be inspected that has a pattern, it is necessary to illuminate the incident direction of the illumination light from an oblique direction with respect to the regular direction of the pattern to average out the effects of unevenness of the pattern and improve the accuracy of focus detection. was there. Further, with conventional optical systems, it has not been possible to simultaneously detect deviations caused by the positional deviation and inclination of the surface to be inspected from the focal point. For this reason, it is recommended to detect only the positional deviation without detecting the tilt of the surface to be inspected and to correct it, or to install a dedicated optical system to detect the tilt. A method has been used in which the inclination of the surface to be examined is first corrected, and when focus detection is performed at individual positions on the surface to be examined, only positional deviations in the perpendicular direction of the surface to be examined are detected and corrected.

It is an object of the present invention to improve accuracy by averaging detection values due to pattern irregularities, and to simultaneously detect positional deviation in the perpendicular direction and deviation due to tilt of the test surface with respect to the focal point of the reduction projection lens. An object of the present invention is to provide a focus detection method and an exposure apparatus using the detection method.

[Means to solve the problem]

In order to achieve the above object, the focus detection method of the present invention focuses a slit (or pinhole, light source, etc.) at a position approximately 20 to 100 mm away from the test surface in terms of optical path length, and This is because the image is formed on a sensor (or a detection slit, etc. as described in the above-mentioned prior art). By doing this, the portion of the illumination light that is narrowly focused (forming a slit image) is separated from the surface to be inspected, so that the cross-sectional area of the light beam irradiated onto the surface to be inspected can be expanded. The cross-sectional area of the light beam irradiated onto the test surface is expanded, illuminating a regular pattern, and the detected values are averaged. In order to be able to detect the deviation with high precision, the slit image formed at a position approximately 20 to 100 ma+ in terms of optical path length from the surface to be inspected moves in proportion to the inclination of the surface to be inspected. Since this slit image is formed on the light-receiving surface of the sensor, it is possible to detect deviations from the focus caused by the inclination of the surface to be inspected. - However, the signal of the slit image detected by this focus detection method contains a combination of the signal caused by the position shift in the perpendicular direction of the surface to be detected and the signal caused by the tilt. By calculating and processing the signals obtained, it is necessary to separate signals caused by the positional deviation in the perpendicular direction of the test surface and signals caused by the inclination of the test surface.

Furthermore, in order to achieve the above object, the exposure apparatus of the present invention is configured such that the light beams forming the first optical path and the second optical path pass through substantially the same optical path in opposite directions. When the optical path is configured in this way, if the surface to be measured shifts (shifts), the light fluxes of the first and second optical paths reflected on the surface to be measured will move in the same direction, as shown in Figure 2. By using the fact that the light beams in the first and second optical paths move in opposite directions when a deviation occurs due to the inclination of the surface to be examined, the surface to be examined is determined by the calculation method described in detail in the Examples. Find the deviation (shift 1-) s with respect to the perpendicular direction and the deviation (pitching) θ, , θ, y due to the inclination of the test surface,
This is to correct the position of the surface to be examined so that it falls within the allowable focal range of the reduction projection lens.

[Effect]

If the image of the slit (or pinhole, light source, etc.) is formed at a position approximately 20 to 100 degrees away from the test surface in terms of optical path length, the cross-sectional area of the luminous flux of the illumination light irradiated onto the test surface will be Because of the magnification, even when detecting a surface to be inspected that has minute irregularities or brightness and darkness, the detected values are averaged and high detection accuracy can be obtained.

At the same time, when the surface to be inspected is tilted, the angle of the reflected light beam changes, and accordingly, the position of the image formed by the slit moves. If a first optical path and a second optical path are provided in which the light beams pass through substantially the same path in opposite directions with respect to the surface to be inspected, the first and second optical paths will be able to compensate for the positional deviation in the perpendicular direction of the surface to be inspected. The slit images formed by the first and second optical paths move in the same direction, and the slit images formed by the first optical path and the second optical path move in different directions in response to a deviation due to the inclination of the surface to be inspected in a direction parallel to the traveling direction of the light beam.

〔Example〕

Embodiments of the present invention will be described with reference to FIGS. 1 to 5. The surface to be inspected 1 (for example, a wafer, etc.) is located at the focal plane of the reduction projection lens 2 that projects a circuit pattern such as an LSI (integrated circuit) onto the surface to be inspected. The optical path for detecting the deviation from the focal plane must enter a narrow region sandwiched between the reduction projection lens 2 and the surface to be examined 1.

The configuration of the optical system according to the present invention will be explained with reference to FIG. The deviation of the test surface l from the focal plane of the reduction projection lens 2 is determined by the inclination (θpx, θpy) of the test surface 1 with respect to the mutually orthogonal directions and the tilt of the test surface 1, as shown in FIGS. It can be broken down into three types of deviations and expressed as positional deviations (s) that are translated in the perpendicular direction. In focus detection according to the present invention, the uniaxial optical system A (or B) shown in FIG. Two types of deviations can be detected: and positional deviation (s). Therefore, in order to be able to detect the inclination (for example, θpy) of the surface to be inspected in the direction perpendicular to the direction of travel of the optical path, the optical system A and the optical path are shown in FIG. 4, which will be described later. The optical system B represented by the dotted line is arranged so that the optical paths constituting each optical system are approximately perpendicular to each other and incident on the test surface 1, and the tilt of the test surface 1 with respect to the focal plane of the reduction projection lens 2 ( Three types of deviations are detected: θ, , θ+py) and positional deviation (s). Since optical system A and optical system B have the same optical path configuration, only the configuration of optical system A will be explained in FIG. In FIGS. 4 and 5, which constitute optical system B, the same parts as those which constitute optical system A are indicated by adding "" to the reference numerals.

Illumination light from an S-polarized light source 3 (for example, a halogen lamp, a light emitting diode, or a semiconductor laser) is focused by a condenser lens 4, and two slits (5a, 5) are used.
b) Light II passes through the slit 5a to form a first optical path using the empty light blocking plate 5, and light II passes through the slit 5b to form a second optical path! Separate at t7. The optical paths 6 and 7, which are S-polarized, pass through the imaging lens 8, are reflected by the beam splinter 9, and are converted into circularly polarized light by passing through the λ/4 phase plate IO. pass through. This optical path 6 is reflected by the test surface 1,
It is reflected by the prism 12, passes through the λ/4 phase plate 10 again, is converted into P-polarized light, and is sent to the beam splitter 9.
passes through the slit 5a to form an image 13a.

The optical path 7 is reflected by the prism 12 and turned back, then reflected by the test surface l, passes through the λ/4 phase plate IO again, is converted into P-polarized light, passes through the beam splitter 9, and passes through the slit 5b. An image is formed 13b.

After passing through the objective lens 14, the slit images 13a and 13b are reflected by the mirror 15 and magnified by the imaging lens 16, and the slit images 13a and 13b are sent to the linear sensor 18, respectively.
13b is imaged. Imaging lens 16 and linear sensor 18
The slit images 13a and 13b are compressed in the longitudinal direction of the slits 5a and 5b perpendicular to the detection direction of the linear sensor 18 by inserting a cylindrical lens 17 between them, and the slit images 13a and 13b are focused on the linear sensor 18. By imaging, the surface to be measured 1 is tilted in a direction perpendicular to the traveling direction of the optical path 6.7 (for example, θ).

7) Linear sensor 1 with slit images 13a and 13b
The movement in the direction perpendicular to the detection direction of 8 is reduced, and even if the surface 1 to be detected is tilted (for example, θ, y), the slit images 13a, 13b will not come off the linear sensor 18, and a good signal can be obtained. is now possible.

FIG. 2 depicts the portion of the optical path 6 that is reflected on the surface to be tested 1 using a luminous flux. The width of the light beam (optical path 6) that has passed through the slit 5b constituting the optical path 6 becomes maximum at the portion where it passes through the imaging lens 8, and converges from there to the slit image 13a. If the light beam between the imaging lens 8 and the slit image 13a is reflected on the test surface l while it is converging, the area of the light beam irradiated on the test surface 1 can be increased, so that minute irregularities can be created on the test surface 1. , even if there is brightness or darkness, the reflected light flux is averaged, and the inclination (for example, θ, y) and positional deviation (s) of the test surface l are
can now be detected with high accuracy. In the opposite direction, optical path 7 stopped at optical path 6 (it can be thought of in the same way).

FIG. 3 shows the principle of detecting the inclination (θI) and positional deviation (s) of the test surface 1 with respect to the focal plane of the reduction projection lens 2 using the optical system A shown in FIG. ■ indicates the case where the test surface 1 is located at the focal plane of the reduction projection lens 2 and there is no inclination (θJ)X) or positional shift (s), with respect to the test surface l. Optical paths 6 and 7 that enter the test surface l from substantially opposite oblique directions travel along optical paths that substantially overlap each other to form slit images 13a and 13b. Slit image 13a11 in this state of ■
The signal waveform that detected 3b is the slope (θ□
) and positional deviation (s).

The deviation of the test surface l from the focal plane of the reduction projection lens 2 that can be detected by the optical system A is determined by the inclination (θ
) and positional deviation (s). Based on the signal waveform in the state shown in (2), the signal waveform of the slit images 13a and 13b from the test surface 1 which has shifted with respect to the focal plane of the reduction projection lens 2 as shown in (2) to (2) is as follows. The positional fluctuations (δ1, δ2) are determined by the method described in , and the positional deviation (s) and inclination (θ□, θ□, θ,y
Note: θ and y are determined by optical system B, which will be described later. This makes it possible to project a fine pattern onto the surface 1 to be inspected. Optical path 6 and optical path 7 when the inclination (θ□) and positional deviation (s) occur simultaneously on the test surface 1 in (2)
The path through which it passes and the slit image 13a created by this
, positional fluctuations from the state shown in ■ in 13b (δ1, δ,
) is shown. In addition, in ■, there is only a positional deviation (s) on the test surface l, and in ■, the passage of the optical paths 6 and 7 when only an inclination (θ□) has occurred on the test surface 1. The respective paths and positional fluctuations (-δ3, δ, , kδ, , δ,) of the slit images 13a and 13b from the state shown in (■) are shown. In (2), if the angle of incidence (θ, ) of the optical paths 6 and 7 on the test surface is sufficiently small, the slit image 13a
, 13b with respect to ■ and the positional shift (s) of the test surface 1 can be expressed by equation (f).

The positional fluctuation (δ3) caused by the positional deviation (s) moves in the same direction for the slit images 13a and 13b, and the amount of movement is almost the same. Since the direction of position variation is reversed, the position variation of the slit image 13a is expressed as -δ. In addition, in the case of ■, the slit image 13 of the optical path 7
The positional variation (δ,) of b is expressed by equation (2) from the inclination (θ□) of the test surface l. When the surface to be tested 1 has an inclination (θI)X), the reflection angle (θ6
) and the reflection angle (θ7) of the optical path 7 with respect to the test surface are expressed by the formula (2,
1) and (2, 2), the light path 6 moves in the opposite direction, but after being reflected on the test surface, it is turned back by the prism 12 and the angular direction is reversed, so the slit image 13a becomes the slit image 13b. move in the same direction. In addition, the test surface l
Since the distance from the reflection to the slit image 13a on optical path 6 is different from optical path 7, a constant (k) is used to correct this optical path length difference.
is determined by equation (3), and the positional variation of the slit image 13a is expressed by 6 degrees. Therefore, as shown in ■, the positional fluctuation (δ,
, δ2) is the positional variation (-53, δ,) of the slit images 13a, 13b when a positional shift (s) occurs on the surface to be inspected 1.
can be expressed by equations (4) and (5), which add the positional fluctuations (kδ, δ,) of the slit images 13a, 13b when the surface 1 to be examined is tilted (θpX). From the above, each positional variation (δ1,
δ2), the inclination (θ,
) can be obtained as follows.

δ, ζ2・S・・・・・・・・・・・・
... (1) Note: If the angle of incidence (θi) of the optical path to the test surface is sufficiently small, δ, ζ2・! 1 ・θm,,l ・・・・・・・・・
・・・・・・・・・(2) θ, = θ, + θ□ ・・・
・・・・・・・・・・・・ (2,1) θ7=θ, -〇,.・・・・・・・・・・・・・・・－(2,2
)51 =k・δ, −δ.

=2-to-j! , −θ,, −2”s +++
(4) δ2; δ, +68 = 2・j! 1 ・θpg +2・S ・−(5
) FIG. 4 shows an example of the overall configuration of an optical system for detecting the focus of the surface to be inspected l as described above. In this optical system, the optical path 6.7 of optical system A described in FIG. Optical system A and optical system B are arranged so that the incident surface is ) is determined by the method described in FIG. 3, and the inclination (θ, ,
) by the same method, it is possible to determine the inclination (θ, yaw, θ, , ) of the test surface 1 with respect to the focal plane of the reduction projection lens 2 and the positional shift (s) in the direction perpendicular to the test surface. I can do it. Although the light source 3 and the light source 3 are provided separately in Fig. 4, it is possible to integrate the light sources of optical system B and optical system B by adding an illumination optical system or branching from the light source with an optical fiber. It is.

A control method for detecting the focus of the surface to be inspected 1 according to the present invention will be explained with reference to FIG. The optical path 6 reflected on the test surface l provided in the optical system A or detected by the fog sensor 18 and the optical path hook (signal waveforms of the slit images 13a and 13b to be imaged are taken into the control device 20, which is the same as this (optical system B) A slit image 13a in which the optical paths 6 and 7'' reflected on the test surface l detected by the linear sensor 18 provided in the
Slit images 13a, 13b are obtained by inputting the signal waveforms of ``, 13b into the control device 20 and detecting the test surface 1 which has no defocus as shown in FIG. and slit image 13a°, 1
Compared with the reference signal waveform of 3b'', the positional fluctuation (δy, δ2) of the slit images 13a, 13b causing a defocus detected by the optical system A and the positional fluctuation (δ, °) of the slit images 13a'', 13b are shown. , δ2).

An example of the signal waveform of the slit images 13a and 13b detected by the linear sensor 18 is shown in the upper right corner of FIG. The signal waveform drawn with two dotted lines is the test surface shown in Figure 3 ■! Tilt to (θ□
) or positional deviation (s) has not occurred, and the signal waveform is a reference signal obtained by detecting the slit images 13a and 13b by the linear sensor 18. In addition, the signal waveform shown by the solid line is inclined (θ
This figure shows an example of a signal waveform detected by the linear sensor 18 when a positional deviation (s-) or ax) occurs, and the difference in position determined from both signal waveforms is the slit image 13a,
13b position fluctuation (δ, δ2). The inclination (θ) and positional deviation (s) of the surface to be inspected 1 are calculated from the values of the determined positional fluctuations (δ, δ2, δ,”, δ2°) using the above-mentioned formulas (ω, ■). .Similarly, position fluctuation (δ1”, δ2°
), calculate the inclination (θ, , ) of the surface to be inspected 1. Based on this signal, the tilt stage 27 on which the test surface 1 is mounted is driven to correct the test surface 1 to within the focus allowed by the reduction projection lens 2.
The signal is converted into a signal necessary for driving 6 and transferred to the driving circuits 21 to 23. Drive circuits 21-23 drive tilt elements 24-26 based on signals received from control device W20.

Tilt elements 24 to 26 are fixed to a stage 28, and drive a tilt stage 27 on which a surface to be examined 1 is mounted to correct the surface to be examined 1 so that it is within the focus allowed by the reduction projection lens 2.

【Effect of the invention〕

It is now possible to accurately detect the deviation of the surface to be inspected (for example, a wafer) from the focus position of the reduction projection lens. It is now possible to transfer patterns.

[Brief explanation of the drawing]

FIG. 1 is a configuration diagram of an optical path that clearly shows the features of the present invention. FIG. 2 shows a portion of the surface to be inspected illuminated by the optical path shown in FIG. 1 using a luminous flux. FIG. 3 shows how the positions of slit images 13a and 13b formed by the reflected light beams change when the optical paths 6 and 7 are irradiated onto a surface to be inspected that is deviated from the focal plane. . FIG. 4 shows the configuration of an optical system for detecting the focus of a reduction exposure apparatus using the focus detection method of the present invention. FIG. 5(6) is a diagram showing a control device for performing main focus detection, and FIG. 5(c) is a diagram showing a signal waveform of a slit image. 1... Test surface, 2... Reduction projection lens, 3... Light source, 4... Condenser lens, 5... Light shielding plate, 8.
1 row: Imaging lens, 9: -beam splitter, l
O...λ/4 phase plate, 12-...prism, 14-...
- Objective lens, 15 - Mirror, 18 - Linear sensor, 20 - Control device, 21 to 23 - Drive circuit,
24-26...Tilt element, 27...Tilt stage, 28... Stage. Agent Patent Attorney Tadashi Akimoto Actual Figure 1 tJ5=l1Cho5B11 111 supporting faces 2° 4-fold shadow K 3: 9F
i4-con light run 12 near-1-414:Mn>ani 15:? Fu 13=9=7beg;7
Figure 2 In+ l 1
Niosho A Shoulder 1 Face 2: Illusion Ho4λ1〉shins=
8 Nio konin shinnu;・9:shi-translation 7・νtsu5 1
〇2 de nii z #1 [] 2” 7.9 sgomu 4th figure 1

Claims (1)

[Claims] 1. A test surface having a pattern on its surface that modulates reflected light, such as minute irregularities and brightness, is illuminated from an oblique direction, and the reflected light from the test surface is received and converted into a signal. In the focus detection method, which determines the deviation of the test surface from the focal plane of a projection lens, etc. that transfers a pattern onto the test surface, this slit image is re-formed by a slit that restricts the passage of illumination light and an optical system. The light is reflected from the test surface in the middle of the convergent optical path, and the position of the slit image formed by the reflected light from the test surface is detected to detect the deviation of the test surface from the focal plane of the projection lens, etc. A focus detection method characterized by calculating. 2. When the illumination light A and the illumination light B limited by the slit are respectively irradiated onto the surface to be inspected from oblique directions substantially opposite to each other, the respective slit images formed by the reflected light A and the reflected light B The positions of A and slit image B move in opposite directions when the surface to be inspected is inclined.
In response to a vertical parallel displacement of the surface to be inspected, the signals move in the same direction and detect the positions of slit image A and slit image B. Taking advantage of the fact that the direction in which B moves is different from the deviation of the test surface, the deviation (s) of the position of equilibrium movement in the perpendicular direction of the test surface and the inclination of the test surface (θ_p_
The focus detection method according to claim 1, characterized in that x) is determined. 3. A linear sensor in which a plurality of fine light-receiving elements are arranged in series is used as the detection sensor, and the positions of the slit image A and the slit image B are determined based on the signal waveform obtained from the linear sensor. The focus detection method according to 1 or 2. 4. A slit image A formed by the reflected light from the test surface;
Displacement (s) of the position parallel to the perpendicular direction of the test surface by calculation processing from the signal obtained by detecting the position of the slit image B
The focus detection method according to claim 1 or 3, characterized in that the value of and the value of the inclination (θ_p_x) of the surface to be inspected are determined respectively. 5. By arranging optical system A and optical system B so that the illumination lights forming optical system A and optical system B are incident on the test surface from directions substantially orthogonal to each other, Displacement of position parallel to the perpendicular direction of the test surface (s)
and the inclination of the test surface in mutually orthogonal directions (θ_p_x, θ
5. The focus detection method according to claim 1, 2, 3, or 4, wherein the focus detection method makes it possible to detect _p_y). 6. Inclination of the test surface with respect to the focal plane (θ_p_x, θ
Based on the signal value due to the positional deviation (s) in the perpendicular direction (_p_y), the position of the test surface is automatically corrected so that it falls within the focus range allowed by the projection lens, etc. that transfers the pattern onto the test surface. 6. The focus detection method according to claim 1, 2, 3, 4, or 5. 7. In an exposure apparatus that sequentially transfers a circuit pattern of a reticle to each shot area made of chips on a wafer through a projection optical system and transfers the circuit pattern to the entire wafer, the reference of the projection optical system of the wafer is used. 6. An exposure apparatus using the focus detection method according to claim 1, wherein a positional shift and a tilt in at least one direction are detected by the same optical system.