JPH0540184A - Stage position detector - Google Patents

Abstract

PURPOSE:To eliminate the detection error due to the air fluctuation and the like in a device making use of an interferometer or the like and obtain a stage position detector with a good detection accuracy. CONSTITUTION:By providing a diffraction grating on a stage 13 which is linearly movable, light beams with different frequencies are guided from different direction so that one side diffraction light of +1 grade and the other side diffraction light of -1 grade are generated perpendicularly to the grating plane. Also provided are a detection means to detect the position of the stage 13 based on the phase difference from a standard signal with a period corresponding to the frequency difference of the two diffraction lights received simultaneously and a control means 12 to control the stage as to have relative velocity to the interference fringes. The moving direction of the interference fringes is changed forwadly and veversely in accordance with the change of the moving direction of the stage 13.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a device for detecting the position of a stage.

[0002]

2. Description of the Related Art A conventional stage position detecting device generally uses a laser light wave interference type length measuring device, a so-called interferometer. This is a fixed point (eg,
The first mirror (fixed mirror) provided inside the laser light source)
And a second mirror (moving mirror) provided on the measured object (stage) whose position is to be measured, and sends light (laser beam) having a known wavelength to both mirrors, The amount of movement of the stage is detected by coaxially synthesizing the reflected light of and receiving the light.

That is, in the conventional interferometer, when the stage moves, the optical path difference is generated between the fixed mirror and the movable mirror by the amount that the movable mirror moves accordingly (the fixed mirror does not move). Since the value of the interferometer (that is, the count value of the counter) changes based on that, the fact that the position of the movable mirror (stage) is obtained from the difference in the optical paths generated here has been used.

[0004]

By the way, with recent technological innovation, for example, a stage apparatus used for an exposure apparatus for semiconductor manufacturing is required to have higher position detection accuracy. A resolution of about 0.01 μm or less is required. However, in position detection using an interferometer, the minimum unit is 1 / of the laser light used.
Since the number of wavelengths is about two wavelengths, the above requirements cannot be satisfied, and a position detection device that replaces this requirement is eagerly desired.

Further, in the position measurement by the conventional interferometer,
Although the distance that light travels and returns is measured using the reflection of a mirror, there is a problem that measurement errors occur due to the effects of air fluctuations in the optical path. That is, if there are areas of air with different temperatures in the optical path of the measurement light, or if there is a portion with a different air density in the optical path due to the effect of wind, etc., the refractive index may change in that optical path. Even when the device itself does not move, there is a problem that the optical optical path changes and the measurement is performed as if the position of the stage was displaced.

Naturally, these errors are included in the measured values as measurement errors when detecting the movement amount of the stage during movement, and in the case of using the conventional interferometer, the errors are accompanied by the movement of the stage. Since the optical path between the fixed mirror and the movable mirror can be long, there is a problem that this error becomes a non-negligible value together with the requirement of the accuracy described above. Further, in the interferometer, since the measurement light beam reciprocates in the optical path, there is a problem that the error of the detection value becomes larger.

Therefore, it is an object of the present invention to provide a new position detecting device capable of reducing the influence of the air fluctuation and improving the measurement accuracy.

[0008]

In order to achieve the above object, according to the present invention, a stage capable of one-dimensional movement in a predetermined direction, a diffraction grating arranged on the stage substantially along the movement direction, and a diffraction grating are provided. Beam irradiating means for irradiating two beams having different frequencies from two directions at a predetermined crossing angle,
Photoelectric detection means for receiving interference light of diffracted lights generated in substantially the same direction from the diffraction grating, and transfer of a detection signal from the photoelectric detection means and a reference signal having a cycle corresponding to the frequency difference between the two beams. Means for detecting the position of the stage based on the difference, and the interference fringes and the stage so that the interference fringes formed on the diffraction grating by the irradiation of the two beams and the stage have a relative speed. There is provided a stage position detection device including a means for controlling at least one moving speed or moving direction.

[0009]

The basic operation of the present invention will be described with reference to the drawings.
As shown in FIG. 5, when a one-dimensional diffraction grating (reflection type or transmission type) is vertically irradiated with a coherent beam, the diffraction angles corresponding to the pitch of the grating pattern are ± 1st order, ± 2nd order,
……… Diffracted light is generated. Conversely, this diffraction grating has a ± 1
When two coherent beams (coherent) are simultaneously irradiated from the generation direction of the second-order diffracted light, ± first-order diffracted light is generated in the direction perpendicular to the diffraction grating.

In FIG. 5, the pitch direction of the diffraction grating WG is the left-right direction of the paper surface, and the two coherent beams (parallel light beams) LB ₁ and LB ₂ are symmetrical at the incident angle θ so that they intersect on the diffraction grating WG. The diffraction grating WG is irradiated with light. In this figure, the beams LB ₁ and LB ₂ are shown by straight lines, but in reality, the beam diameter (cross-sectional area) is sufficient to irradiate the diffraction grating WG uniformly.

When the beams LB ₁ and LB ₂ are emitted from the same laser light source, a one-dimensional interference fringe IF is formed on the diffraction grating WG. This interference fringe IF
Is a bright and dark fringe parallel to the diffraction grating WG, and its pitch P _f is defined by the relationship of the following expression (1). (Λ is the wavelength of the beams LB ₁ and LB _2. )

[0012]

[Equation 1]

Now, let us consider the diffracted light generated from the diffraction grating WG by the irradiation of the _one beam LB ₁ . Generally, the diffraction angle A _n of the ± n-order diffracted light generated when one beam is irradiated to the grating having the pitch P _g is _expressed by the following equation (2) with reference to the 0-th order light (regular reflection light or transmitted light). It becomes a relationship.

[0014]

[Equation 2]

As shown in FIG. 5, since the incident angle of the beam LB ₁ is θ, the regular reflection light of the beam LB _{1 on} the diffraction grating WG, that is, the emission angle of the 0th order light B ₁ (0) is also θ. Is.
Then, with the 0th-order light B ₁ (0) as a reference, when clockwise in the plane of the drawing is set to positive (+) and counterclockwise is set to negative (-),
By irradiation of the beam LB ₁ generated from the diffraction grating WG +
The 1st-order light B ₁ (+1) and the -1st-order light B ₁ (-1) are located symmetrically with the 0th-order light B ₁ (0) interposed therebetween. Here, when the −1st order light B ₁ (−1) is generated perpendicularly to the diffraction grating WG, the 0th order light B ₁ (0) and the −1st order light B ₁ (−1) are generated.
It is necessary to set the angle formed by and, that is, the diffraction angle A _n at n = 1 to θ. Therefore, from equation (2), n = 1, A _n = θ
Then, the following relationship is established.

[0016]

[Equation 3]

Substituting this into equation (1), the following relationship holds.

[0018]

[Equation 4]

As is clear from the equation (3), the pitch P _g of the diffraction grating WG can be calculated by changing the incident angle θ (sin θ = λ / P
_g ) the pitch P of the interference fringes IF uniquely determined by
_{If it} is exactly doubled from _f , the minus _first -order light beam B ₁ of the beam LB ₁
(-1) is generated vertically from the diffraction grating WG. +1
The next light B ₁ (+1) is generated in a direction spreading outward by an angle θ from the 0th light B ₁ (0). Similarly, of the ± first-order diffracted lights B ₂ (+1) and B ₂ (−1) generated from the diffraction grating WG by the irradiation of the other beam LB ₂ , the + 1st-order light B ₂
(+1) is generated vertically from the diffraction grating WG. And
Similarly, the −1st order light B ₂ (−1) is generated in a direction spreading outward by an angle θ from the 0th order light B ₂ (0).

Therefore, when the pitch P _f of the interference fringes IF is 1/2 of the pitch P _g of the diffraction grating WG, ± 1st order lights B ₁ (−1) and B ₂ (+1) of the same order and opposite signs are used. Are combined coaxially (direction perpendicular to the diffraction grating surface) and interfere with each other. Moreover, the diffracted lights that interfere with each other are ± 1
In the case of FIG. 5, not only the next light, but also the diffracted lights having the order difference of 2 are combined coaxially, so that they also interfere with each other. For example, the + second-order light B ₂ (+2) generated from the diffraction grating WG by the irradiation of the beam LB ₂ is coaxial with the 0th-order light B ₁ (0), and the + 1st-order light B ₁ (+1).
Coincident with is the + 3rd order light B ₂ (+3).

By the way, in the position measurement, as shown in FIG. 5, diffracted lights (diffracted lights of order difference 2) generated in substantially the same direction and interfering with each other are photoelectrically detected. Here, B ₁ (−1) and B ₂ of the ± first-order light in FIG.
It is assumed that the interference light of (+1) is photoelectrically detected. First, when the diffraction grating WG in FIG. 5 is moved in the pitch direction (x direction), the light phases of the _two beams LB ₁ and LB ₂ change as follows.

2 · π · x / P _g , −2 · π · x / P _g
(Unit is rad)

Therefore, the interference intensity of the ± 1st order light is cos.
The pitch P _{g of the} diffraction grating WG changes with (4πx / P _g ).
A sinusoidal photoelectric signal having one cycle is obtained for each 1/2 of the displacement. Therefore, based on the intensity change of the photoelectric signal when the diffraction grating WG and the interference fringes IF are relatively moved,
The displacement of the diffraction grating is measured.

The above description is based on the two beams LB ₁ and LB.
_When the frequencies of ₂ are equal and the two beams LB ₁ ,
If there is a frequency difference Δf in LB ₂ , the interference fringes IF on the grating will run (flow) in the pitch direction at a velocity V. The moving speed V of this interference fringe is represented by V = P _f · Δf.

Therefore, assuming that the frequency of the beam LB ₁ is f ₁ and the frequency of the beam LB ₂ is f ₂ , the frequency difference Δf = f
₁ -f _2, and the phase of light when the beam LB _1, LB ₂ in FIG. 5 is as follows even kept stationary diffraction grating WG.

[0026]

[Equation 5]

Therefore, the ± first-order light B ₁ (-
The interference intensity I _t of 1) and B ₂ (+1) is expressed by the following equation.

[0028]

[Equation 6]

Since a time component is included in this equation (6), the interference intensity temporally changes at the frequency of Δf (= f ₁ -f ₂ ), and its phase component is (2x / P). _g ) means that location information is included. And this △
f is called a beat frequency, and in order to measure the phase (2x / P _g ), the same beat frequency (reference signal) serving as a reference is generated, and the phase difference from it is measured to measure the position information (reference The amount of positional deviation with respect to the position) can be obtained.

In accordance with the above explanation of the principle, the present invention adopts a method of measuring the positional deviation of the diffraction grating by giving the frequency difference Δf to the _two beams LB ₁ and LB ₂ , and providing the diffraction grating directly on the moving stage. Therefore, the stage position is detected. That is, the present invention is based on the above-described basic principle and irradiates two light fluxes having slightly different frequencies on a grating pattern provided on a stage at a predetermined angle, and, for example, interference light between ± first-order diffracted light beams generated here. The position information is detected based on the beat signal obtained by photoelectrically detecting, and the position information of the movable one-dimensional stage or the like is known using this position information.

For this reason, in the present invention, the diffraction grating (which has a grating pattern) is provided on the one-dimensionally movable stage, and its pitch direction is substantially the same as the moving direction. It is installed in. And
Two light fluxes having slightly different frequencies are made to enter the lattice pattern from the irradiation means from two different directions. In this case, the angle (2θ) sandwiched by the two light beams in the optical means is set so that one first-order diffracted light and the other −1st-order diffracted light are generated in the same direction (sin θ = λ / P _g is satisfied). At this time, the diffracted lights having order differences of 2 other than the ± first-order lights are diffracted in the same direction as described above. Here, due to problems such as the intensity of diffracted light and the arrangement of the detection system,
It is preferable to determine the incident angles of the two light fluxes so that the first-order diffracted light and the other -1st-order diffracted light are generated in a direction orthogonal to the measurement direction (direction perpendicular to the diffraction grating surface).

Further, as a result of irradiating the grating pattern with the two light fluxes by the detection means, the photoelectric sensor receives the diffracted light having the second order difference of 2 (for example, ± 1st order diffracted light), and changes the brightness of the interference fringes. Cycle (ie beat frequency △
A sinusoidal AC signal corresponding to f), that is, a so-called beat signal (including position information) is detected. At this time, a reference grid pattern having the same interval (pitch) as that provided on the movable stage is engraved at a fixed position outside the stage, and an irradiation device by the same light source means as described above is used for this grid pattern. Is incident, the ± 1st-order diffracted light generated there is received by a photoelectric sensor, and the beat signal detected here is used as a reference signal. Then, the position of the stage is measured by detecting the phase difference between the reference signal received here and the measurement signal from the grating pattern provided on the moving stage.

As the reference diffraction grating, the one having the same pitch (P _f = P _g ) provided on the stage was used, and the 0th order light and the ± 1st order light generated from the reference diffraction grating were used. It may be a reference beat signal.

By the way, the position detecting device according to the present invention uses a diffraction grating unlike the conventional one using a mirror when an interferometer is used. The moving direction of the stage, which is detected when detecting the position, is perpendicular to the mirror surface in the conventional interferometer, whereas it is parallel to the diffraction grating surface in the present invention. That is, the interferometer irradiates the inspection light in a direction perpendicular to the mirror provided on the surface to be detected, and detects the position of the stage that moves in the same direction as the irradiation direction. On the other hand, in the present invention, with respect to the diffraction gratings arranged substantially along the moving direction of the stage, the irradiation direction of the inspection beam is changed by the beam irradiation means and the photoelectric detection means provided at the position facing the grating surface. However, the position detection and the movement amount are detected from the direction parallel to the moving direction of the stage. Therefore, the optical path length of the measurement light does not change due to the movement of the stage, and since it can be detected at a position close to the diffraction grating surface, the influence of air fluctuations is extremely small, and the accuracy is extremely high. Position measurement is possible.

Further, the width of the two light beams incident on the diffraction grating is widened (that is, the beam system is thickened) to increase the overlapping portion (intersection region) of the beams of the two light beams, that is, the direction orthogonal to the measurement direction (non-). By creating a sufficiently long intersection area in the measurement direction), the range to be measured becomes wider, and the so-called depth of focus of the measurement light can be deepened. this is,
As shown in FIG. 4, according to the present invention, the two light beams LB ₁ and LB ₂ are inclined to the lattice pattern 409 provided on the stage 403 by an angle θ with respect to the perpendicular 405 formed from the lattice pattern forming surface of the stage 403. Irradiate. Here, when each beam diameter is “a”, the intersecting area is a vertical line 4
There is a distance of "a / sin θ" in the 05 direction. Therefore,
Since position measurement can be performed if there is an object to be measured in this intersection region, the present invention has a large depth of focus in the direction to be measured with respect to the stage 403 (object to be measured). Therefore, the measurement error is reduced, and the detection accuracy is improved.

By the way, one wavelength (360 °) of the interference fringes formed on the diffraction grating is 1 / of the grating pattern pitch.
It is 2. Therefore, the resolution of this one wavelength is expressed by the following equation.

[0037]

[Equation 7]

Here, the lattice pattern pitch P = 8 μm,
When the resolution d of the phase difference meter is 0.1 °, the resolution is 1.1 nm, and it can be seen that the present invention provides a remarkably large resolution as compared with the conventional interferometer.

Next, as described above, the pitch P _{I of the} interference fringes formed on the diffraction grating is defined by the relationship expressed by the following equation.

[0040]

[Equation 8]

Here, the above-mentioned relational expression of the incident angle sin θ
=? / P, it turns out that P _I = P / 2 results. This means that the pitch of the interference fringes is P / 2, so the speed V at which the fringes flow at this time is
It is expressed by the following formula.

[0042]

[Equation 9]

Therefore, as is apparent from this equation, the velocity V of the interference fringes generated by changing the frequencies of the two incident light beams can be changed.

Here, in the present invention, since the direction in which the interference fringes flow is the pitch direction of the diffraction grating, it is the same as the moving direction of the stage (regardless of forward and backward directions). Therefore, when the moving speed of the stripe and the moving speed of the stage become the same (the direction and the speed are the same), there arises a problem that the beat signal is not generated. In the present invention, in order to compensate for this, control is performed so that the stage and the interference fringe always have a relative velocity so that these velocities do not become the same.

As an example of control by this control means,
It is conceivable to set the frequencies of the two incident light beams in advance so that the speed of the interference fringes is faster than the maximum moving speed of the stage. In addition, the incidence 2 depending on the moving speed of the stage
A method of changing the frequency of the light beam or conversely limiting the moving speed of the stage so as not to exceed the moving speed of the interference fringes is conceivable, but any method may be adopted.

When this stage is a so-called stepper used for semiconductor manufacturing, etc., the exposure time (stopping time) and the subsequent moving direction (direction) and speed are known in advance, so that the stepper moves. Direction
In accordance with the above, since the mutual movement speeds of the interference fringes are opposite to each other, the mutual movement speeds always have a relative speed, so that it is not necessary to control the movement speed. The direction of movement of the interference fringes can be changed only by changing the set frequencies of the two light beams.

Here, in order to change the moving speed of the interference fringes,
It suffices to change one of the frequencies of the measurement light beam (change the frequency difference). For example, by changing the modulation condition (high frequency signal) of the acousto-optic modulator (AOM) that performs frequency modulation of the measurement light, The frequency difference changes and the moving speed changes. At this time, the reflection direction of the diffracted light may change, but since it is only a slight frequency change, it does not deviate from the pupil of the light receiving portion of the detection means. However, in the case of making a drastic frequency change, the incident angle to the diffraction grating may be changed accordingly. Further, in order to make the moving directions of the interference fringes normal and reverse, the AOM of the two light fluxes is used.
It suffices to carry out change control in which the modulation conditions of (2) are switched to each other and the frequencies of the two light fluxes are switched. An optical waveguide may be used instead of the AOM.

[0048]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A stage position detecting device according to an embodiment of the present invention applied to a two-dimensionally movable stage will be described with reference to the drawings. Here, even if the stage is two-dimensionally movable, two stages that are one-dimensionally movable, for example, an X stage that moves only in the X direction are installed on a Y stage that moves only in the Y direction. And
Needless to say, if it is a so-called XY two-dimensional moving stage, it is possible to detect the two-dimensional position of the stage position by incorporating the position detecting device of the present invention into each stage.

First, as shown in FIG. 3, the X stage 33
The diffraction grating pattern 39a having a pitch in the moving direction is formed on the side surface along the moving direction of a, and similarly, the diffraction grating pattern 39b is also formed on the side surface of the Y stage 33b.
Then, individual detection systems (not shown) are provided so as to face these individual diffraction grating patterns. At this time,
The detection system in the X direction is fixed to the Y stage 33b, and the detection system in the Y direction is fixed to the base of the Y stage. That is, the detection system is fixed to a portion where the distance does not change in the vertical direction of the diffraction grating surface.

Here, in order to simplify the description, the present embodiment will be described in more detail with reference to FIG. 1, taking only the position detection of the X stage as an example. The X stage 13 that is movable in the X direction by the motor 14 or the like is provided with a lattice pattern 19 with a constant interval (pitch: P) on the side surface along the X direction. Aiming at the lattice pattern 19, si
It saw the conditions of nθ = λ / P, so that one plus two light beams coming back to the vertical-order diffracted light and the other of the -1st-order diffracted light to the grating pattern is interference, frequency f ₁ and frequency f _Two light beams LB ₁ and LB ₂ having a value of ₂ are made incident.

An example of this light source means will be described with reference to FIGS. 1 and 2. An example of this type of light source means is also disclosed in, for example, Japanese Patent Application Laid-Open No. 2-272305. In summary, the light flux emitted from the light source 1 is split into two light fluxes by the two-flux frequency shifter 2 and both are modulated by an optical frequency modulator (AOM) to have two frequencies having known frequencies f ₁ and f _2. The light beams LB ₁ and LB ₂ are emitted.

First, the schematic structure and operation of the two-beam frequency shifter 2 will be described with reference to FIG. He-N
A linearly polarized (eg, P-polarized) laser beam (parallel light flux) from the light source 1 made of an e-laser is emitted by the quarter wavelength plate 20.
After passing through 1 to be converted into circularly polarized light, it passes through a convex lens 202 and enters a polarization beam splitter (PBS) 203, where it is divided into a P-polarized component and an S-polarized component.

The PBS 203 makes the incident beam orthogonal to each other.
P polarization beam LB ₁ is PB
The S-polarized beam LB ₂ is transmitted through S203 and is incident on the acousto-optic modulator (AOM) 206. The S-polarized beam LB ₂ is reflected by the PBS 203, and is transmitted through the mirror 204 to the acousto-optic modulator (AOM) 2.
It is incident on 05. The laser device and the AOMs 205 and 206 that form the light source 1 do not necessarily have to be on the same plane as the two-dimensional stage.

Here, the AOM 206 is the drive frequency S
The AOM 205 is modulated with the high frequency signal of F ₁ , and the AOM 205 is modulated with the high frequency signal of the drive frequency SF ₂ . At this time, for example, the frequency SF ₁ is 80 MHz and the frequency SF ₂ is 79.
It is set to 95 MHz. In addition, AOM205, 2
When 06 is driven, high-order diffracted light is generated in addition to the 0th-order light L ₀ . Of these, the first-order diffracted lights LB ₁ and LB
Only ₂ is taken out through the slits 207 and 208.

The frequency of the _first-order diffracted light LB ₁ (LB ₂ ) is fl, which is the original frequency of the light of the He--Ne laser beam.
Letting (light flux / wavelength) be fl + SF ₁ (SF ₂ ). Therefore, the frequency difference Δf between the beams LB ₁ and LB ₂ extracted from the slits 207 and 208 is (fl
_{+ SF 1) - (fl +} SF 2) = SF 1 -SF 2 = 50
It becomes KHz.

Here, by changing the modulation condition (drive frequency) in the AOMs 205 and 206, both frequencies can be changed and the speed at which the interference fringes formed on the diffraction grating flow can be changed. Therefore, in this embodiment, the AOM
The voltage applied to 205 and 206 can be changed arbitrarily. Further, since the moving speed of the interference fringes is changed according to the change of the moving speed of the stage, the control device 12 (FIG. 1) is used. AOM20 based on control signal
The voltage applied to 5,206 is variable.

Further, parallel plane glasses 210 and 211 are provided in both optical paths, and by tilting them, the position of the optical path (that is, half-plane beam splitter 21 to be described later) is provided.
The distance between the beams LB ₁ and LB ₂ emitted from the beam _No. 3) can be changed, and the angle of incidence θ on the diffraction grating can be changed. Further, as shown in FIG. 2, the light emitted from the AOM 205 passes through the slit 207 and then enters the half-wave plate 211. Therefore, A
The S-polarized beam LB ₁ emitted from the OM 205 is 1/2
It is converted into P-polarized light by the action of the wave plate 21 and coincides with the polarization direction of the beam LB ₂ emitted from the AOM 206.

Then, the beam LB ₁ is incident on the half-plane beam splitter (HBS) 213 via the mirror 212, and the S-polarized beam LB ₂ (frequency f ₂ ) is directly transmitted to the HBS.
It is incident on 213. Here, as shown in detail in FIG. 6, the HBS 213 has a total reflection mirror MA on one half of the bonding surface.
Is vapor-deposited, and the beam LB ₁ is made incident thereon to be reflected with a light amount of almost 100%. On the other hand, the beam LB ₂ passes through the transparent portion of the joint surface as it is. As a result, the principal rays of the _two beams LB ₁ and LB ₂ become parallel to each other and are positioned symmetrically with respect to the optical axis of the position measurement optical system.
The beams become book beams LB ₁ and LB ₂ .

As described above, the two light beams LB ₁ and LB ₂ emitted from the two light beam frequency shifter 2 are HB so that their principal rays are parallel to each other with the optical axis of the position detection optical system interposed therebetween.
It is incident on S3. Here, the beam LB ₁ has a frequency f ₁ ,
The beam LB ₂ has a frequency f ₂ , and the frequency difference Δf is set to 50 KHz as described above. Further, these polarization directions are aligned in the same polarization direction, for example, P polarization, but a ¼ wavelength plate (not shown) is provided before entering the PBS 3 so that they are circularly polarized with each other. It is aligned with. Therefore, the polarization beam splitter (P
The beam LB ₁ incident on BS) 3 is wavefront-divided into a P-polarized beam LB _1P and an S-polarized beam LB _1S having a frequency f ₁ , and a beam LB ₂ is a P-polarized beam LB _2P and an S-polarized beam LB having a frequency f _2. Wavefront is divided into _2S and.

Here, the two P-polarized beams LB _1P (frequency f ₁ ) and LB _2P (frequency f ₂ ) that have passed through the PBS ₃
Intersect with each other on a reference diffraction grating 5 fixed at any position on the device. At this time, the reference diffraction grating 5
Are fixed via a lens system (inverse Fourier transform lens) 4 for converting the pupil into an image plane, and are arranged in the focal plane of the lens system 4. Then, the beams LB _1P and LB _2P become parallel light fluxes that enter the reference diffraction grating 5 from two different directions at a predetermined crossing angle, and they cross each other on the reference diffraction grating 5 and are transmitted. Therefore, the interference fringes IF 'are formed on the reference diffraction grating 5, and the interference fringes IF' move at the velocity V '.

Then, the reference grating 5 and the beams LB _1P and LB
The pitch relationship with the interference fringes created by _2P is P _f '= P _g '
Stipulated in. Therefore, as shown in detail in FIG. 7, the 0th-order light (solid line) of the beam LB _1P transmitted through the reference grating 5,
The + 1st order light (broken line) of the beam LB _1P generated from the reference grating 5 is combined in the same direction to form the interference light BT (0, + 1).
And the 0th order light of the beam LB _2P (solid line) and the beam LB
The -1st-order light (broken line) of _1P is combined in the same direction to receive the interference light BT (0, + 1), BT (0, -1) on the same light receiving surface, and the reference signal SR of frequency Δf (25 KHz) Is output. The reference signal SR is input to the phase difference detector 10 and gives the reference value of the measurement signal.

On the other hand, the two S-polarized beams LB _1S (frequency f1) and LB _2S (frequency f2) reflected by the PBS3.
Enters the objective lens 7. 2 through the objective lens 7
The book beams LB _1S and LB _2S irradiate the grating mark Gw on the stage 13 at incident angles θ and −θ and intersect each other.
Interference fringes flowing at a velocity V in the pitch direction are formed on the lattice mark Gw, and the pitch P _{f of the} interference fringes is determined by the lattice mark G.
It is set to 1/2 of the pitch P _{g of} w. Therefore, as described with reference to FIG. 5, the interference light BT (± 1) of ± first-order light is vertically generated from the wafer W. Interference light BT (± 1)
Since the incident beams LB _1S and LB _2S are both parallel light fluxes, they are similarly generated as parallel light fluxes, travel backward through the objective lens 7, and are reflected by the mirror 8 to reach the photoelectric element 9. still,
The photoelectric element 9 is arranged in a substantially pupil plane of the position detection optical system (here,
It is arranged at the front focus position of the objective lens 7.

In addition, the photoelectric element 9 has an interference light BT (±
The zero-order light components BT (0, + 2) and BT (0, -2) of the beams LB _1S and LB _2S are also incident side by side on both sides of 1). Here, the photoelectric element 9 is composed of a plurality of light receiving elements PD _{1 to} PD ₃ as shown in FIG.
The central element PD ₁ is used. In FIG. 8, the broken line represents the pupil conjugate plane Ep '. The interference light BT (± 1) and the interference lights BT (+2,0) and BT (0, −) shown in FIG.
In 2), both converge on a minute spot (1 mm or less) on the pupil plane and are separated from each other.

Although only the photoelectric signal (SG described later) from the element PD ₁ is used here, for example, the signal strengths of the _three elements PD ₁ to PD ₃ are compared and the signal strength is the highest. You may make it use the thing. Of course, only one relatively large light receiving surface is provided in the photoelectric element 9, and a spatial filter is arranged in front of the photoelectric element 9 so that light other than the interference light BT (± 1) does not enter the light receiving surface. It is also possible to use only BT (± 1).

Then, the intensity change of the interference light BT (± 1) is photoelectrically converted by the light receiving element PD ₁ , and the measurement signal S
It is output as G. The same applies to the position detection system for the Y direction.

Here, the interference light B detected by the photoelectric element 9
Since T (± 1) is intensity-modulated with frequency Δf,
The signal SG is an AC signal having a frequency Δf (50 KHz). The output signal SG of the photoelectric element 9 corresponds to the above equation (6), and this signal SG is the photoelectric element 6 for reference.
It is input to the phase difference detector 10 together with the signal SR from.
Then, the phase difference detector 10 obtains the phase difference Δφ between the detection signal SG and the reference signal SR, and calculates the corresponding positional deviation amount Δx (or Δy) of the lattice mark Gw by the following equation.

[0067]

[Equation 10]

The position detecting operation will be described in more detail. In the phase difference detector 10 shown in FIG. 1, the magnitude of the reference signal SR and the magnitude of the measurement signal SG are set to 500 KHz as shown in FIG. Digital sampling is performed and at least one cycle waveform of each signal (10 sampling points) is stored in a memory. Then, the following calculation is performed with the waveform of the signal SR as Fref. And the waveform of the signal SG as Fsig.

[0069]

[Equation 11]

[0070]

[Equation 12]

Here, sin θ and cos θ are standard waveform data stored in advance in the memory, and θre
f. and .theta.sig. represent the phases of the signals SR and SG, respectively.
Therefore, the phase difference detector 10 further includes the phases θref.
The difference from g. θsig.−θref. = Δφ is calculated, and the positional deviation amount Δx (or Δy) is obtained from the above equation (7). Further, when both phases are shifted by one wavelength, the number is stored in the counter 11, and the positional deviation amount Δx is measured within one wavelength.

Then, main controller 12 calculates the calculated shift amount Δx (or Δy) and the information on the moving distance of wafer stage 13 stored in counter 11 (the number of phases shifted by one wavelength). The current position of the wafer stage 13 is determined based on At this time, when changing the moving speed of the stage or the moving measure of the interference fringes, the control device 12 issues a control signal to the motor 14 or the two-beam frequency shifter 2.

As a premise of starting the position measurement, the counter 11 is reset at the reference position mark provided at the edge of the lattice mark or in the vicinity thereof, and at the same time, the reference deviation amount Δx ′ at the reference position is taken in as the offset amount. Then, the subsequent position detection is calculated based on the preset value at the reference mark (relative position detection from the reference position is performed). In addition, when changing the moving direction (direction) of the interference fringes, the position at the time of changing is stored in memory, and the count value and position shift amount in the opposite direction are calculated backwards to obtain the preset value (reference position). mark)
Calculate the relative position from.

As described above, according to this embodiment, the phase of the diffracted light returning from the grating pattern of the moving stage changes according to the position of the stage, and the change amount Δφ moves Δx of the stage. On the other hand, they have the following relationship.

[0075]

[Equation 13]

For this reason, the + 1st-order diffracted light and the -1st-order diffracted light undergo phase changes of opposite signs, and the phase change of the beat signal from the detector 9 is as follows.

[0077]

[Equation 14]

Therefore, the device can detect the minute displacement amount Δx of the stage by detecting Δφ. In addition, if the control system counts the number (shifted by one wavelength) when capturing the point (position shifted by one wavelength) and the direction where the phase difference becomes 0, the displacement amount of a long distance can also be obtained. Easy to grasp.

The diffracted light emitted when the grating pattern is irradiated with two light fluxes also has a higher order, so ± 1.
The beat signal can be obtained not only by the next-order light but also by using these higher-order diffracted lights (a combination of those having a difference in order of 2, for example, 0th-order light and + second-order light). Further, as shown in the system of the reference lattice 5, it is possible to apply the one having the order difference of 1 instead of the one having the order difference of 2.

Further, in the present embodiment, an example in which the present invention is applied to an XY two-dimensional moving stage has been described, but a Z stage is provided on this and the moving direction (Z direction, XY in FIG. 3).
By providing a grid pattern along the direction (perpendicular to the plane), X
It can also be applied to a YZ three-dimensional moving stage. In addition, by providing two sets of detection systems in the X direction of the stage, the yawing of the stage can be corrected. Note that the half-plane beam splitter 213 in FIG. 2 may be a mere beam splitter, and in this case, there is an advantage that the PBS 3 becomes unnecessary. Furthermore, grid pattern 1
The two light fluxes for irradiating 9 are, for example, P-polarized light and S-polarized light,
The analyzer may be arranged in front of the photoelectric element 9.

[0081]

As described above, according to the present invention, two luminous fluxes having different frequencies are sent to the grating pattern provided on the stage, and based on the beat signal detected from the diffracted light returning in substantially the same direction. Since the stage position is known, the optical path lengths of the first light flux and the second light flux are constant and short, so that the effects of air fluctuations and the like are reduced.

Further, since the movement amount is detected by the so-called heterodyne method, there is an advantage that the resolution is high and even a minute change amount can be detected, and since the phase difference is detected, it is strong against the fluctuation of the light amount. There are advantages.

Further, when it is known in advance that the moving speed of the stage exceeds the speed of the interference fringes, or when the moving time is long, that is, when it is known that the moving speed increases, Δ
By increasing the initial setting of f (the difference between the frequencies of the two modulated light fluxes), it is possible to prevent the stage and the interference fringes from always having a relative velocity, which makes it impossible to detect the position. However, if the moving speed of the interference fringes is too fast (the frequency difference is large), the resolution will decrease, so it must be determined in consideration of the resolution. When the decrease in resolution is a problem, such a problem can be avoided by controlling the moving speed of the stage or controlling the moving direction of the interference fringes in the opposite direction.

On the other hand, when the present invention is applied to a stepper, since the movement direction of the stage (stepper) is programmed in advance by the movement control means of the stepper, the interference fringes always move in the opposite direction to the movement direction of the stage. As described above, the above problem can be solved by controlling the frequency of the incident light to be reversed.

[Brief description of drawings]

FIG. 1 is an explanatory diagram showing a schematic configuration of a stage position detection device according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram showing a detailed configuration of a two-beam frequency shifter in the above embodiment.

FIG. 3 is a configuration diagram showing a stage and a lattice mark provided therein in the embodiment.

FIG. 4 is an explanatory diagram illustrating a state in which two light beams intersect with each other (a region where the two light beams overlap) in the detection system of the present invention.

FIG. 5 is an explanatory diagram for explaining how a diffraction grating of the present invention is irradiated with two light beams to obtain a beat signal from the diffracted light.

FIG. 6 is a configuration diagram illustrating a half-plane beam splitter of the two-beam frequency shifter in the above embodiment.

FIG. 7 is a configuration diagram illustrating a reference signal detection system in the above embodiment.

FIG. 8 is a configuration diagram illustrating a light receiving element for detecting a beat signal in the above embodiment.

FIG. 9 is an explanatory diagram illustrating a method of detecting a positional deviation amount at the time of position detection in the above embodiment.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Laser 2 ... Two-beam frequency shifter 13, 33, 403 ... Movable stage 6, 9 ... Detector (photoelectric element) 205, 206 ... AOM (optical frequency modulator) 39a, 39b ... Lattice pattern 10 ... Phase difference detector 11 ... Counter 12 ... Control device 14 ... Motor

Claims (1)

[Claims] 1. A stage capable of moving one-dimensionally in a predetermined direction; a diffraction grating arranged on the stage substantially along the movement direction thereof; and 2 having different frequencies from two directions at a predetermined crossing angle on the diffraction grating. Beam irradiation means for irradiating one beam; photoelectric detection means for receiving interference light of diffracted light generated in the same direction from the diffraction grating; detection signal from the photoelectric detection means and frequency difference between the two beams Means for detecting the position of the stage based on a difference in transfer with a reference signal having a cycle corresponding to the above; and an interference fringe formed on the diffraction grating by irradiation of the two beams and a relative speed of the stage. A stage position detection device, comprising: means for controlling the moving speed or moving direction of at least one of the interference fringes and the stage.