JP3085248B2 - Position detecting device and position detecting method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a high-resolution displacement detection device using a diffraction grating, and more particularly to a displacement amount of a diffraction grating formed on a wafer or a mask, or an alignment error between a mask and a wafer. The present invention relates to a device for detecting

[0002]

2. Description of the Related Art Conventionally, as this kind of displacement detecting device, Japanese Patent Application Laid-Open Nos. 61-208220 and 61-2284 have been disclosed.
As disclosed in Japanese Patent Publication No. 49, a diffraction grating is formed on a substrate such as a wafer, and two coherent light beams are irradiated from different directions on the diffraction grating to form interference fringes arranged in the pitch direction of the diffraction grating. A method of detecting the displacement amount and the displacement amount of the diffraction grating with reference to the static interference fringes is known. Furthermore, in the above publication, when the edge of each grating element of the diffraction grating on the substrate is accompanied by asymmetry, the intensities of various orders of the diffracted light from the diffraction grating are individually obtained in advance and made into a database,
It is also disclosed that an error due to asymmetry is corrected based on a database at the time of detecting an actual position shift.

Further, a method of detecting an amount of misalignment of a diffraction grating by an optical heterodyne method in which an interference fringe applied to the diffraction grating is caused to flow at a high speed in one direction is disclosed in, for example, JP-A-61-2.
No. 15,905, and JP-A-62-56818. According to the technique disclosed in this publication, a constant frequency difference is given to two coherent light beams, and the two light beams intersect at a predetermined angle on a diffraction grating, so that the two light beams can move in one direction at a speed corresponding to the frequency difference. A flowing interference fringe can be created. In detecting the amount of displacement, the interference intensity of diffracted light beams of the same order generated from the diffraction grating in the same direction is photoelectrically detected, and a measurement photoelectric signal having a frequency equal to the frequency difference between the two light beams is obtained. On the other hand, a reference signal having a frequency equal to the frequency difference between the two light beams is separately prepared, and the phase difference of the diffraction grating in the pitch direction is detected by detecting the phase difference of the measurement photoelectric signal with respect to this reference signal. It is measured within ± 1/4 pitch.

Further, such an optical heterodyne type position shift amount detecting apparatus is provided with a TTR (through the reticle) in a projection type exposure apparatus (stepper) for manufacturing a semiconductor device.
The one incorporated as an alignment system is disclosed in
No. 283129. The TTR alignment system is usually used in a state where one shot area on a wafer is aligned with a projected image of a reticle pattern at about ± 1 μm, and a grid mark of the shot area is detected and a displacement of 1 μm or less is detected. It is used to finely move the wafer stage or the reticle stage so that the deviation amount is obtained and corrected. Thus, TTR
In the alignment system, the two light beams and the wafer (grid mark) do not move relative to each other when the grid mark is detected, but instead the interference fringes on the grid mark are always flowing, so that the phase difference measurement can be sequentially performed.

[0005]

When a wafer is measured as an object to be measured, a resist layer is usually applied to the entire surface of the wafer in the exposure step. Therefore, interference fringes 2
The light beam irradiates the diffraction grating mark on the wafer surface via the resist layer. The resist layer is a thin film having a thickness of about 1 μm. On the other hand, the laser beam or the like having a good monochromaticity is used for the two light beams, so that a thin film interference phenomenon or the like easily occurs. The intensity of the diffracted light (for example, ± first-order light) may change extremely. For wafers that have undergone a series of processes,
The duty (ratio of line width to space width) of the grid mark itself may change from 50%, and also in this case, the intensity of the diffracted light (± first-order light) to be detected changes.

Further, if both the resist layer and the grating duty work in a bad direction, the intensity of the diffracted light to be detected (interference light of ± first order light) is hardly obtained. . Therefore, an object of the present invention is to provide a device that can detect the displacement of a diffraction grating even when the above-described adverse effects occur.

[0007]

According to the present invention, there is provided a light beam irradiating means for irradiating a diffraction grating provided on an object to be measured with light beams from two different directions. Photoelectric detecting means for detecting the interference intensity of specific diffracted light beams generated from the diffraction grating, and measuring means for measuring a displacement of the object to be measured in a grating pitch direction of the diffraction grating based on an output signal of the photoelectric detecting means; And a first photoelectric detector for detecting the interference intensity of diffracted light beams having the same diffraction angle traveling in the same direction from the diffraction grating, and a different diffraction angle traveling in the same direction from the diffraction grating. A second photoelectric detector for detecting the interference intensity of the diffracted light beams having the following. The first measuring unit measures a value corresponding to the displacement of the object to be measured based on the output signal of the first photoelectric detector. Measuring unit and A second measuring unit for measuring based on an output signal of the second photoelectric detector, wherein the light beam irradiating means includes: a frequency modulation member for giving a predetermined frequency difference to the light beams from the two directions; A reference signal generator for outputting a reference signal having an equal frequency, whereby each of the first photoelectric detector and the second photoelectric detector has a first photoelectric signal and a second photoelectric signal having a frequency equal to a predetermined frequency difference. A first phase detector for detecting the first phase relationship between the reference signal and the first photoelectric signal as a measured value, and the second measuring unit outputs the reference signal and the second signal. A second phase detection unit that detects a second phase relationship with the photoelectric signal as a measurement value,
The measuring means includes an offset selection unit that gives a predetermined offset to the detection result of the second phase difference detection unit based on the detection result of the first phase detection unit and the detection result of the second phase detection unit. Further, according to the present invention, in a light beam irradiating means for irradiating a light beam to the diffraction mark provided on the object to be measured, and in an apparatus for measuring the displacement of the object to be measured based on optical information generated from the diffraction mark, A first photoelectric detector for detecting first diffracted light information generated from the diffraction mark; a second photoelectric detector for detecting second diffracted light information generated from the diffraction mark; information from the first photoelectric detector; (2) An arithmetic unit having a plurality of modes for calculating a value corresponding to the position of the device under test using at least one of information from the photoelectric detector; and a selecting unit for selecting a plurality of modes. And

As in the prior art, the basic structure of the present invention is as follows: a light beam irradiating means for irradiating the diffraction grating with a light beam from two directions; and a photoelectric detecting means for detecting the interference intensity of diffracted light generated from the diffraction grating. Measuring means for measuring the displacement of the diffraction grating in the pitch direction based on the output signal. Further, as a novel configuration of the present invention, two photoelectric detectors, a first photoelectric detector and a second photoelectric detector, are provided.

The first photoelectric detector detects the interference intensity of diffracted light beams of the same diffraction angle generated from the diffraction grating by irradiation of the two beams, and the second photoelectric detector detects the diffracted light beams of different diffraction angles generated from the diffraction grating. Detect the mutual interference strength. Further, in the present invention, a first measuring section and a second measuring section are provided as measuring means. The first measurement unit measures a value corresponding to a displacement amount (that is, a displacement amount) of the diffraction grating (object to be measured) based on an output signal (a phase difference from a reference signal) of the first photoelectric detector, and The second measurement unit measures a value corresponding to the displacement amount (position shift amount) of the diffraction grating based on the output signal (phase difference from the reference signal) of the second photoelectric detector. And it has decided to have the offset selection part which calculates the offset given to the measurement value of a 2nd measurement part based on the measurement value of a 1st measurement part, and the measurement value of a 2nd measurement part.

[0010]

The first photoelectric detector of the present invention detects the interference intensity of diffracted light beams having the same diffraction angle. The first photoelectric detector detects two light beams that irradiate the diffraction grating from two directions by a first beam and a second beam. In this case, for example, the incidence angle and grating pitch of the two light beams on the diffraction grating are set so that the -1st-order diffraction light generated by the irradiation of the first beam and the + 1st-order diffraction light generated by the irradiation of the second beam travel in the same direction. Is achieved by defining the relationship. If this condition is satisfied, for example, the 0th-order diffracted light (specular reflection light or linearly transmitted light) generated by the irradiation of the first beam and the + 2nd-order diffracted light generated by the irradiation of the second beam travel in the same direction. This means that this is detected by the second photoelectric detector as interference intensity.
The present invention pays attention to the fact that the probability that the condition in which the first-order diffracted light is weakened and the condition in which the second-order diffracted light is weakened at the same time due to the influence of the duty of the diffraction grating, the resist layer, and the like is extremely small is extremely low. If a sufficient light intensity is obtained in the photoelectric detector, the displacement (positional displacement) of the diffraction grating is detected using one of the output signals. Therefore, usually, position shift detection is performed based on the interference intensity of diffracted light beams of the same order (± 1st order), and when the interference intensity is below a certain level or cannot be obtained at all, a different order (0 order, 2 order)
By switching so as to detect the displacement based on the interference intensity between the diffracted lights in the next step, the detection rate of the diffraction grating mark is reliably improved.

Here, the principle of the present invention will be described with reference to FIGS. In FIG. 2, two light beams BM ₁ and BM ₂ are symmetrically applied to the reflection type diffraction grating GA at an incident angle ψ. The first beam BM ₁ and the second beam BM ₂ are
A parallel light beam is emitted from the same laser light source and has a sectional area sufficient to irradiate the entire surface of the grating GA. Here, the frequencies of the first beam BM ₁ and the second beam BM ₂ are assumed to be equal. However, even if a frequency difference is provided, the following description is applied in the same manner.

Incidentally, the frequency difference is,
This is the beat frequency of the in-type
It is easily provided by using the light source. Also the first beam
BM ₁And the second beam BM_TwoFrequency is different from
Does not necessarily mean that the wavelengths are different.
Rather, they may be considered substantially equal. ,
Heterodyne beat frequency is lower than radio frequency
(10^Three-10^FiveHz), which is, for example, He-Ne
Absolute frequency of 628 nm typical wavelength of laser light
(Light speed / wavelength) about 4.76 × 10¹⁴Hz
Clearly, it can be ignored.

Now, as shown in FIG. 2, two beams B
When M ₁ and BM ₂ are irradiated, an interference fringe IF is formed in the intersection area. The interference fringes IF are set so as to be parallel to the grating elements of the grating GA. Where two beams B
Let the wavelengths of M ₁ and BM _{2 be} λ and the grating GA (duty 50
%) Is Pg and the pitch of the interference fringes IF is Pf, first, the following relationship is established.

Pf = λ / 2 sinψ (1) Further, assuming that each n-th order diffracted light of the first beam BM ₁ and the second beam BM ₂ is generated perpendicular to the surface of the grating GA, The following relationship holds. sinψ = n · λ / Pg (2) Therefore, from Expressions (1) and (2), 2 · n · Pf = Pg. In FIG. 2, n = 1 is shown, and the sign (positive or negative) of each diffraction order is positive for higher-order diffracted light that spreads clockwise with respect to 0-order diffracted light, and negative for counterclockwise. .

First, the zero-order diffracted light B ₁ (0) generated by the irradiation of the first beam BM ₁ travels in the same optical path as the second beam BM ₂ . And it caused by the irradiation of the first beam BM ₁ -1 order diffracted light B ₁ (-1) proceeds vertically, -2 order diffracted light B ₁ (-2) is backward substantially the same optical path as the first beam BM ₁ . Further, the + _{1st order} diffracted light B ₁ (+1) generated by the irradiation of the first beam BM ₁ is the 0th order diffracted light BM
_From (1), the beam advances in a direction inclined approximately clockwise by an angle ψ, and the -3rd-order diffracted light B ₁ (-3) advances from the first beam BM _{1 in} a direction inclined approximately angle ほ ぼ in a counterclockwise direction. on the other hand,
Caused by irradiation of the second beam BM ₂ 0-order diffracted light B ₂
(0) is backward the same optical path as the first beam BM _1. Then, the + 1st-order diffracted light B ₂ (+1) generated by the irradiation of the second beam BM ₂ advances vertically, and the + 2nd-order diffracted light B ₂ (+2)
It will reverse the substantially same optical path as the second beam BM _2. Further, the −1st-order diffracted light B generated by the irradiation of the second beam BM ₂
M ₂ (−1) travels from the 0th-order diffracted light B ₂ (0) in a direction counterclockwise inclined substantially by an angle ψ, and the + 3rd-order diffracted light B
₂ (+3) travels in a direction inclined clockwise approximately by an angle か ら from the second beam BM ₂ . Therefore, ± 1st order diffracted light B ₁
Under the condition that (-1) and B ₂ (+1) are generated in the same direction, the 0th-order diffracted light B ₁ (0) and the + _2nd-order diffracted light B ₂ (+
2) is in the same direction, and the zero-order diffracted light B ₂ (0) and −2
The next-order diffracted light B ₁ (−2) is in the same direction. Plus +1
The _first-order diffracted light B ₁ (+1) and the + 3rd-order diffracted light B ₂ (+3) are in the same direction, and the −1st-order diffracted light B ₂ (−1) and the −th-order diffracted light B ₁ (−3) are in the same direction. .

The diffracted lights traveling in the same direction interfere with each other, and the interference intensity changes sinusoidally due to the relative movement of the interference fringe IF and the grating GA. FIG. 3 (A),
(B) shows the state of generation of each diffracted light when the _two beams BM ₁ and BM ₂ are separately irradiated on the grating GA. The vertical axis indicates the relative intensity of the diffracted light, and the horizontal axis indicates It represents the diffraction angle of each diffracted light with the normal direction of the grating GA as the zero reference.

The + 2nd-order diffracted light B ₁ (+
2), + 3rd order diffracted light B ₁ (+3) and − in FIG.
Second order diffracted light B ₂ (-2), third order diffracted light B ₂ (-3)
Are not shown in FIG. As is apparent from FIG. 3, the absolute value of the difference between the order of the diffracted light generated by the first beam BM ₁ and the order of the diffracted light generated by the second beam BM ₂ becomes 2, and the two are the same. Proceed in the direction and interfere with each other. That is, when the ± n-order diffracted lights are in the same direction, the diffracted lights of the order in which the absolute value of the difference is 2n travel in the same direction.

According to the present invention, a plurality of interference light beams of the order diffracted light whose absolute value of the difference is 2n are individually photoelectrically detected, and it is determined which output signal is to be used for the positional deviation detection. The detection accuracy is improved by giving an offset to the light.

[0019]

FIG. 1 shows a configuration of a displacement detecting apparatus according to a first embodiment of the present invention. In this embodiment, the displacement of a diffraction grating Gw formed on a semiconductor wafer W is expressed by ± 1/1 of the grating pitch Pg.
Consider the case of detection within four. The basic configuration of the optical system shown in FIG. 1 is the same as that disclosed in, for example, JP-A-63-283129. Two parallel beams B ₁ ,
B ₂ is linearly polarized light in the same direction, has a frequency difference Δf (for example, 100 KHz or less), and is condensed by the lens system 10. The beams B ₁ and B ₂ are incident on the wafer W by the polarization-neutral beam splitter 12.
M ₁ , BM ₂ and beams Br ₁ , Br ₂ incident on the reference system
And divided into The beams BM ₁ and BM ₂ further pass through the polarizing beam splitter 14 at almost 100%, enter the quarter-wave plate 16, and enter a telecentric objective lens system 18.
, And intersect as a parallel light beam on the wafer W.

The rear focal plane of the lens system 10 coincides with the front focal plane of the objective lens system 18, that is, the pupil plane ep,
The two beams BM ₁ and BM ₂ are converged on a minimum spot at a symmetrical point across the optical axis in the pupil plane ep. Beam B
The principal rays of M ₁ and BM ₂ are almost parallel to the optical axis in the pupil space (between the lens system 10 and the objective lens system 18).
Therefore, the beams BM ₁ and BM ₂ that have been emitted from the objective lens system 18 and directed to the wafer W both become parallel light beams and have an incident angle 傾 that is symmetrically inclined with respect to each other. In this embodiment, since the wafer W is arranged so as to coincide with the rear focal plane of the objective lens system 18, its focal length is F, and the distance from the optical axis of the beam spot in the pupil plane ep is l. Then, the incident angle ψ is represented by tan ψ = 1 / F (3).

The beams BM ₁ and BM ₂ that have passed through the polarization beam splitter 14 are, for example, linear P-polarized light.
When the light is incident on the wafer W by the wave plate 16, it becomes circularly polarized light whose polarization plane is rotated in the same direction and interferes with each other. Therefore, interference fringes having a pitch Pf flow on the diffraction grating Gw at a constant speed, and ± first-order diffracted light returns along the optical axis of the objective lens system 18 perpendicularly to the wafer W as described with reference to FIG. The ± 1st-order diffracted lights from the grating Gw become interference lights BT (± 1) that interfere with each other, pass through a quarter-wave plate 16, where they are converted from circularly polarized light into linear S-polarized light. Almost 100% is reflected and received by the photoelectric element 20a. The photoelectric element 20a is an objective lens system 1
8 on the optical axis at or near the pupil plane. Since both the ± first-order diffracted lights from the grating Gw of the wafer W are parallel luminous fluxes, the interference light BT (± 1) is focused to the smallest spot size at the position of the photoelectric element 20a.

Similarly, the beams BM ₁ , BM
_The interference light BT (−
2,0) and BT (0, + 2) are also converted from circularly polarized light to S-polarized light by the quarter-wave plate 16 and
And are received by the photoelectric elements 20b and 20c, respectively. The photoelectric elements 20b and 20c are arranged on or near the pupil plane of the objective lens system 18 so as to sandwich the photoelectric element 20a.

On the other hand, the two beams Br ₁ and Br ₂ split by the beam splitter 12 are converted by the lens system 22 into parallel light beams that intersect on the reference grating 24. The front focal plane of the lens system 22 coincides with the rear focal plane of the lens system 10, and the rear focal plane of the lens system 22 has a reference grating 2
4 matches. The reference grating 24 is of a transmission type, and the 0th-order light therefrom is blocked by an aperture 26, and only the ± 1st-order diffracted light BR (± 1) traveling coaxially is received by the photoelectric element 28 through the aperture 26.

Two beams Br intersecting at the reference grating 24
_Since both ₁ and Br ₂ are linear P-polarized light, they interfere with each other and form interference fringes flowing on the reference grating 24.
The reference grating 24 serves as a reference point on the apparatus side when detecting a positional shift, and is stably fixed in the apparatus. Now, since the original beams B ₁ and B ₂ have a frequency difference Δf, the interference light BR (± 1) from the reference grating 24 and the grating G
BT (± 1), BT (-2,0), BT
In (0, +2), the intensity changes in a sinusoidal manner at the beat frequency Δf. Therefore, the photoelectric elements 20a, 20b, 20
If c and 28 have a responsiveness that sufficiently follows the beat frequency Δf, each photoelectric element will have a frequency Δf
Output the sine wave signals Sa, Sb, Sc, and SR.

The output signal Sa has a phase difference detector 40 for detecting a phase difference Δθ ₁ between the output signal Sa and a reference signal, and an intensity detector for detecting a signal strength or amplitude GSa of the output signal Sa. And 41.
The output signals Sb and Sc are output from a phase difference detection unit 42 for detecting a phase difference Δθ ₂ from the output signal SR, the intensity of the signals Sb and Sc,
Alternatively, the values GSb and GSc relating to the amplitude are input to the intensity detection unit 43 which individually detects the values. The phase difference detection unit 42 individually obtains the phase difference Δθb between the output signal Sb and the reference signal SR and the phase difference Δθc between the output signal Sc and the reference signal SR, and then calculates the phase difference Δθ _{2 by the} following equation. Ask for.

Δθ ₂ = (Δθb + Δθc) / 2 (4) The intensity determination unit 44 determines whether the interference light BT (−2,0), (0, +
The values GSb and GSc relating to the intensity in 2) are input, and the following expression is calculated as an example, and the average intensity GU is obtained. GU = (GSb + GSc) / 2 (5) Further, the intensity determination unit 44 sends the determination output Q (a, to the displacement calculating unit 45 based on the magnitude relationship between the intensity value GSa and GU.
b) and outputs an error signal ERR when neither of the intensity values GSa, GU satisfies a predetermined condition.
Is output.

The displacement calculator 45 calculates the phase differences Δθ ₁ , Δθ
₂ and the reference grating 2 of the diffraction grating Gw is calculated by the following equation.
4, the amount of displacement (displacement) ΔX with respect to the grid pitch Pg
Calculate within ± 1/4 of Δθ = (a · Δθ ₁ + b · Δθ ₂ ) / a + b (6) ΔX = {(Pg / 2) / 2π｝ · Δθ …………………… (7) , 0 ≦ a ≦ 1, 0 ≦ b ≦ 1) The coefficients a and b in the equation (6) are set by the judgment output Q (a, b) from the intensity judgment unit 44. A judgment algorithm is provided.

Therefore, three typical algorithms incorporated in the strength judging section 44 will be described below. In equation (7), -π <Δθ <π, and therefore -P
g / 4 <ΔX <Pg / 4. The first algorithm IF GSa> GU THEN a = 1 , b = 0 IF GSa <GU THEN a = 0, b = 1 second algorithm preset signal strength value with g _0, IF GSa> g ₀ THEN a = 1, b = 0 IF GSa <g ₀ THEN a = 0, b = 1 Third algorithm a = C ₁ · GSa, b = C ₂ · GU (C ₁ and C ₂ are constants) The algorithm 1 uses the intensity of the interference light BT (± 1) between the first-order diffracted lights and the interference light BT of the second-order and zero-order diffracted lights.
By comparing the average intensity of (−2, 0) and (0, +2),
In this method, the phase difference Δθ and the positional deviation amount ΔX are obtained based on only the signal having the larger intensity. The second algorithm basically uses the signal Sa (phase difference Δθ ₁ ) of the interference light BT (± 1), and only when the intensity value GSa falls below a certain level g ₀ , the signals Sb and Sc (positions). This is a method using the phase difference Δθ ₂ ). The third algorithm is the intensity value GS
In this method, a specific gravity is given to each of the phase differences Δθ ₁ and Δθ ₂ according to a and GU, and both phase differences Δθ ₁ and Δθ ₂ are always used. These three representative algorithms are appropriately selected depending on changes in the duty ratio of the diffraction grating Gw on the wafer W, the thickness of the resist layer, and the like, and the most accurate algorithm is used. Actually, the state of the lattice mark can be checked in advance by visual observation with a microscope or a television,
The accuracy is confirmed based on the result of trial grilling after aligning the pieces. As described above, in the configuration of the present embodiment,
The intensity determination unit 44 determines two interference lights (−2, 0), (0, +
In the case where the difference between the intensity values GSb and GSc in 2) is extremely large and one of the intensity values is almost zero, and the intensity value GSa of the output signal Sa is insufficient for detecting the displacement. , An error signal ERR is output to warn the operator.

It is desirable that the intensity detectors 41 and 43 detect the amplitude of the signal. For example, if the diffracted light beams of the same order are the same, as long as the intensities of the _two irradiation beams BM ₁ and BM ₂ are substantially equal, the intensity is almost the same. Therefore, the modulation of the signal amplitude is ideally 100%. Become. However, since the diffracted light beams differ from each other, the original intensity is theoretically different as shown in FIG. 3, so that the modulation degree of the signal amplitude does not become 100%.

FIG. 4 schematically shows the degree of modulation of the signal amplitude. FIG. 4A shows, for example, a signal Sa corresponding to a temporal change in the intensity of the interference light BT (± 1).
FIG. 4B shows an example of the waveform of the interference light BT (−2,
It is an example of a waveform of a signal Sb (or Sc) corresponding to a temporal intensity change of (0), (0, +2). FIG. 4 (A),
In (B), V ₁ and V ₂ are amplitudes. In the case of interference light BT (± 1) of ± 1st-order diffracted light, the modulation degree (amplitude / average value,
Or AC / DC) is large, and the degree of modulation in the case of 0th-order and 2nd-order interference light is small. However, the degree of modulation or the absolute value of the amplitude changes sensitively depending on the shape of the grating mark Gw, the influence of the resist layer, and the like. Therefore, position shift detection is usually performed using interference light of the same order diffracted light generated at a diffraction angle symmetrical to the optical axis, and when the amplitude of the intensity change of the interference light decreases with the modulation factor. Are interference lights BT (−2, 0) of diffracted lights of different orders, (0, +
2), and if the amplitude is sufficient, switch to use the interference light, that is, the second
It is desirable to give priority to the algorithm.

Incidentally, the phase difference detecting section 42 in FIG.
Although the phase differences Δθ _b and Δθ _c of the output signals Sb and Sc with respect to the respective reference signals SR are obtained, and the simple average value Δθ ₂ is output, inconvenience occurs in practice.
The phase difference of the interference light BT (± 1) of the ± first-order diffracted light with respect to the reference signal SR is Δθ ₁ , and the phase difference Δθ
Let's focus on the phase differences Δθ _b-1 and Δθ _c-1 between _b and Δθ _c .

Here, Δθ _b-1 = Δθ _b -Δθ ₁ , Δθ
_c-1 = Δθ _c −Δθ ₁ and the phase difference Δ
θ _b-1 and Δθ _c-1 are −180 ° to + depending on the step, duty, resist thickness and the like of the lattice mark Gw on the wafer W.
It can vary between 180 °. However, if each grid element of the grid mark Gw is made substantially symmetrical in the position shift detection direction (the generation direction of the diffracted light), then Δθ _b-1 ≒ −Δθ _c-1
It is. Therefore, for example, the displacement is zero (Δθ ₁ = 0 °)
In this case, it is assumed that there is a lattice mark that should satisfy Δθ _b-1 = + 170 ° and Δθ _c-1 = −170 °. At this time, it is assumed that the misalignment of the lattice mark has happened, and the measurement has been performed with Δθ ₁ = + 30 °.

FIG. 5 is a vector diagram showing this state. In FIG. 5, the clockwise direction is positive and the counterclockwise direction is negative.
In this case, the correct values of the phase differences Δθ _b and Δθ _c are Δθ _b
= + 200 °, Δθ _c = −140 °, and Δθ ₂ =
(Δθ _b + Δθ _c ) / 2 = 30 ° = Δθ ₁ However, in general, the phase difference can be detected only from -180 ° to + 180 °, so that Δθ _b = + 200 ° is Δθ _b = −1
It will be measured as 60 °. Therefore, when the phase difference Δθ ₂ is obtained by simple averaging, Δθ ₂ = (Δθ _b + Δθ
_c ) / 2 = (− 160−140) / 2 = −150 °, which is a value completely different from Δθ ₁ = 30 °.

However, assuming that Δθ _b-1 ≒ −Δθ _c-1 is assumed, the phase difference detector 42 can correct the phase difference. For this purpose, the internal configuration of the phase difference detecting section 42 is, for example, as shown in FIG. Although the configuration of FIG. 6 is actually made by an arithmetic program such as a computer, it is schematically shown here in the form of a circuit block. In FIG. 6, the signal Sb of the interference light BT (−2, 0) and the interference light BT (0, +2)
Are input to the phase difference measuring units 100 and 102, and the phase differences Δθ _b and Δθ _c from the reference signal SR are obtained. The phase difference measuring units 100 and 102 may be either hardware processing or software processing using a waveform memory, and detect a phase difference within ± 180 °.

The phase differences Δθ _b and Δθ _c are respectively _calculated by the subtractor 1
04, 106, the subtractor 104 calculates Δθ _b and Δθ ₁
The calculated difference [Delta] [theta] _b-1 of the (output of phase difference detector 40), the subtracter 106 obtains a difference [Delta] [theta] _c-1 of the [Delta] [theta] _c and [Delta] [theta] _1.
The condition determination unit 108 determines that the difference Δθ _{b- 1} and Δθ _c-1 are Δθ
It is determined whether or not the condition of _b-1 ≒ -Δθ _c-1 is satisfied, and one of the three output data of the offset selection unit 110 is determined based on the result. Osefoot selection unit 110 outputs 0 °, +360
° or -360 ° is sent to the adder 112. The adder 112 adds the phase difference Δθ _b and the offset value and outputs the result to the averaging unit 114. Averaging unit 114
Is the offset phase difference Δθ _b and the phase difference measurement unit 1
Adding the phase difference [Delta] [theta] _c from 02 obtains an average value, ±
The phase difference Δθ ₂ when the interference light between the second-order diffracted light and the zero-order diffracted light is used is output. Thus, the operation of the block in FIG. 6 will be described using the numerical values exemplified above.

First, as shown in FIG. 5, when Δθ ₁ = 30 °, the phase difference Δθ measured by the phase difference
_b is −160 ° (correctly + 200 °), and the phase difference Δθ _c measured by the phase difference measuring unit 102 is −14.
0 °. Therefore, the output value Δθ _b-1 of the subtractor 104 is -190 °, and the output value Δθ _c-1 of the subtractor 106 is -170.
°. The condition determining unit 108 has, for example, a function of adding Δθ _b-1 and Δθ _c-1 . If the added value is almost zero, it is determined that Δθ _b-1 ≒ Δθ _c-1 is true. I do. Here, Δθ _b-1 + Δθ _c-1 ≒ −360 °, which is a negative value.
It is set to output an offset value of 360 °.

As a result, the adder 112 is set at + 360 °
Is added to the phase difference Δθ _b (−160 °) to output a correct value of + 200 °.
Note that when the calculation result of Δθ _b-1 + Δθ _c-1 is almost zero, the offset value is 0 °, and when it is a positive value, the offset value is selected to be −360 °. Further, the adder 112 may be configured to add the output side of the phase difference measurement unit 102, that is, the phase difference Δθ _c to the offset value. In that case, Δθ _c (−
140 °) and the offset value + 360 ° is added to +2
20 °, but the averaging unit 114 similarly calculates Δθ ₂ ≒
It is calculated as 30 °.

From the above, Δθ ₁ = (Δθ _b + Δ
θ _c ) / 2 is an ideal case. If such an ideal state is satisfied in most cases, the configuration as in the present invention would not be necessary. However, in actual wafer processing, usually Δθ ₁
Since ≠ (Δθ _b + Δθ _c ) / 2, the method shown in FIG. 6 is not always successful.

For example, in the above numerical example, Δθ _b (−16
0 °) and Δθ _c (−140 °) as they are, and if only Δθ ₁ is detected as −150 ° (correctly + 30 °) due to an accidental extreme error, then Δθ _b−1 = −10 °, Δ
θ _c-1 = + 10 °, which satisfies the condition of Δθ _b-1 ≒ −Δθ _c-1 , so that an accurate displacement cannot be obtained.

However, such an extremely large error
Because the probability of a difference occurring is very low,
Modulo Δθ_b, Δθ_cIt is effective to add a correction to.
FIG. 7 shows two beams B in the apparatus of FIG.₁, B_TwoSupply
This shows an example of a light transmission system.
It is assumed that a laser light source is used. The laser light source 50
The laser beam LB is converted into a beam LB by the beam splitter 52.₁,
LB_TwoBeam LB₁Is reflected by the mirror 53
Through a parallel plane glass (plain parallel) 55
Into the frequency shifter (acousto-optic modulator: AOM) 57
Shoot. The other beam LB_TwoIs plane parallel 5
4 and is incident on a frequency shifter (AOM) 56.
AOMs 57 and 56 are a high-frequency oscillator 70 and a frequency synthesizer
Driven by high-frequency drive signal made by Iser 72 etc.
And the driving frequency f for the AOM 57₁Is, for example, 8
0.0 MHz and the driving frequency f for AOM56
_TwoIs, for example, 80.050 MHz. AOM57,
56 is an incident beam LB₁, LB_Two0th order light and higher order diffracted light
And emits only the first-order diffracted light.
B₁, B_TwoApertures 58 and 59 to be taken out as
Is provided.

Beam B transmitted through aperture 59₁Lap of
The wave number represents the frequency of the original laser beam LB as f₀Then f₀
+ F₁become. Beam B that has been made parallel light so far₁Is
6
3. The beam is split into the pupil conjugate plane ep 'via the beam splitter 65.
The light is collected as a pot light. On the other hand, the aperture 58 is transparent.
Beam B passed_TwoIs f ₀+ F_TwoBecomes Len
Convergence by the system 62, the plane parallel 64,
Spots on the pupil conjugate plane ep 'via the beam splitter 65
It becomes light and condenses.

Here, the two beams B ₁ and B ₂ are not coaxially synthesized by the beam splitter 65, but the lens system 6 is used.
The eccentricity is combined such that the principal ray of the beam is positioned symmetrically with respect to the optical axis of No. 6 and parallel to the optical axis. The front focal point of the lens system 66 coincides with the pupil conjugate plane ep ′, and the two beams B ₁ and B ₂ are placed on a parallelogram aperture (illumination field stop) 67 disposed at the rear focal point of the lens system 66. And the parallel light flux intersects.

The position of the aperture 67 coincides with the front focal point of the lens system 10 in FIG. Therefore, the pupil conjugate plane ep ′ is conjugate with the pupil plane ep in FIG. Here, the frequency difference between the _two beams B ₁ and B ₂ is (f ₀ + f
_{2) - (f 0 + f} 1) = 80.05MHz-80.0M
Hz = 50 KHz, which is the beat frequency of the heterodyne system. With the above configuration, plane parallel 5
The beam LB ₁ , LB ₂ is changed by changing the inclination of 4, 55.
Can be adjusted optimally with respect to the AOMs 56 and 57, and by changing the inclination of the plane parallels 63 and 64, the two beams B ₁ and B ₂ on the pupil conjugate plane ep ′ or the pupil plane ep (1 in FIG. 1), that is, the intersection angle and incident angle of the two beams on the object plane and the image plane can be adjusted. The aperture 67 is formed in a parallelogram so that an opening edge extending in a direction perpendicular to the grating pitch direction of the diffraction grating Gw is not formed. This is because the influence of the diffraction (Fran-Hofer diffraction) by the edge of the aperture 67 is received. This is to avoid reaching the system.

By the way, when a light transmitting system as shown in FIG. 7 is used.
In this case, the other beam exit surface of the beam splitter 65
1 in the direction of arrow AA,
The reference grating 24, the aperture 26, and the photoelectric element 28 are arranged.
And the beam splitter 12 may be replaced with a total reflection mirror.
No. As described above, using the two AOMs 56 and 67
Note frequency Δf (f₁−f _TwoThe way to create
Since the frequency Δf can be changed freely, the response of each photoelectric element
The optimum beat frequency according to the character is obtained. In addition
Two beams with a frequency difference, such as a
System can be prevented from being coaxial.
Noise components due to necessary stray light, crosstalk, etc.
Generation of an AC signal with a wave number) is extremely small
There are advantages.

Next, a second embodiment of the present invention will be described with reference to FIG. 8. FIG. 8 is a modification of the light transmitting system shown in FIG. 8, the apertures 58 and 59, the lens systems 61 and 62, and the plane parallels 54, 55, 63, and 6 in FIG.
4 are similarly provided, but are not shown.

The linearly polarized laser beam LB changes its polarization plane substantially.
After entering the half-wave plate 51 rotated by 45 °,
Polarizing beam splitter PBS₁P-polarized beam and S-polarized light
Split into beams. P-polarized beam circulates around AOM56
Wave number f₁Is shifted only by beam B_TwoBecome a synthetic pre
It reaches the total reflection surface MR of the mechanism 65A. On the other hand, S-polarized
Is reflected by the mirror 53 and enters the AOM 57, where
Frequency f_TwoIs shifted only by beam B₁And the synthesis
Transmits the rhythm 65A. Synthetic prism 65A is a beam
B₁, B_TwoAre combined so that the chief rays of
I do. Beam B ₁, B_TwoRotates the plane of polarization by approximately 45 °
Polarizing beam splitter PBS through half-wave plate 68
_TwoIncident on. Polarizing beam splitter PBS_TwoIs S bias
Light beam B₁To the S-polarized beam Br₁And P polarized beam BM
₁Into a P-polarized beam B_TwoThe S-polarized bee
M Br_TwoAnd P polarized beam BM_TwoVector decomposition into
These beams Br₁, Br_TwoIs the reference system (the lens in FIG. 1)
System 22 and later), beam BM₁, BM_TwoIs detected
Go to the system (after the polarization beam splitter 14 in FIG. 1)
No.

With such a light transmission system configuration, the utilization efficiency of the beam is higher than that of FIG. 7, and the light amount ratio of the _two beams B ₁ and B ₂ can be correctly adjusted to 1: 1. There is an advantage that the light amount ratio between the beams Br ₁ and Br ₂ going to the detection system and the beams BM ₁ and BM ₂ going to the detection system can be freely set. First, in order to adjust the intensity ratio between the beams B ₁ and B ₂ , the 板 wavelength plate 51 may be slightly rotated around an axis parallel to the laser beam LB. This way,
Light intensity ratio of beams Br ₁ and Br ₂ , beams BM ₁ and BM ₂
Of the reference grating 24 and the diffraction grating G
The intensity of the two beams illuminating w can be made exactly equal. Further, when the half-wave plate 68 is slightly rotated around the optical axis, the beams Br ₁ and Br ₂ and the beam B
The relative division ratio between M ₁ and BM ₂ can be changed,
From the viewpoint of signal detection, a small amount of light may be sent to a stable reference system, and a large amount of light may be sent to a detection system for an object to be measured such as a wafer W. Thus, the example of FIG. 8 has an effect that the intensity of each beam can be easily adjusted.

FIG. 9 shows a modification of the light transmitting system and the light receiving system according to the third embodiment of the present invention. The configuration before the combining prism 65A is the same as that of FIG. P polarized beam B ₁ (f ₀ + f
₁ ) and the S-polarized beam B ₂ (f ₀ + f ₂ ) are split by a polarization-neutral beam splitter BS ₃ , one of which becomes a P-polarized beam Br ₁ and an S-polarized beam Br ₂ to form a lens system 22. And is converted into parallel light beams that cross each other. The reference grating 24 is arranged at the intersection, but since the beams Br ₁ and Br ₂ have complementary polarizations, no interference fringe IF occurs. ± 1st-order diffracted light diffracted by the reference grating 24 and traveling in the same direction as the optical axis (here, no interference)
Is an interference light for the first time when it becomes an analyzer 27 through an aperture 26, and a photoelectric element 2 that outputs a reference signal SR
Reach 8.

On the other hand, the P-polarized beam BM ₁ and the S-polarized beam BM ₂ transmitted through the beam splitter BS ₃ are combined with the lens system 10, the mirror 11 and the neutral beam splitter BS.
_The light enters the objective lens system 18 via ₄ . The objective lens system 18 converts the _two beams BM ₁ and BM ₂ into parallel luminous fluxes and intersects with the grating Gw on the wafer W. Again beam BM
_Since the polarization directions of ₁ and BM ₂ are different, no interference fringes occur on the grating Gw. ± 1 vertically generated from the lattice Gw
The next diffracted light returns along the optical axis of the objective lens system 18, is reflected by the beam splitter BS ₄ , and reaches the photoelectric device 20 a via the analyzer 19. Similarly 0-order, 2-order diffracted light is also reflected by the beam splitter BS _4, the photoelectric element 20b through the analyzer 19, it reaches 20c. Photoelectric element 20a, 2
Reference numerals 0b and 20c have light receiving surfaces individually formed on the same semiconductor substrate, and individually output output signals Sa, Sb and Sc. The light receiving surface is arranged so as to be coincident (or conjugate) with the pupil plane of the objective lens system 18. Also in this light receiving system, by providing the analyzer 19, the ± first-order diffracted lights or the 0th-order and second-order diffracted lights can interfere with each other, and the output signals Sa, Sb, Sc can be obtained at the same beat frequency. it can.

As shown in FIG. 9, two AOMs 56, 6
If the S-polarized beam and the P-polarized beam are separately frequency-shifted at 7 and the two beams are decentered and combined by the combining prism 65A, the coaxial of the original two frequencies is used as in the case of using a Zeeman laser light source. There is an advantage that the S / N ratio is remarkably improved as compared with the case where the beam is split into two beams.

Next, the embodiment of FIG. 4 of the present invention will be described with reference to FIGS.
1 will be described. In the fourth embodiment, the pitch of the diffraction grating Gw of the wafer W is set to の of that in the first embodiment, while maintaining the incident angle の of the _two beams BM ₁ and BM ₂ as it is. The pitch of the grating 24 is also halved. FIG. 10 schematically shows the state of generation of diffracted light when the pitch of the diffraction grating Gw is halved. When the crossing angle of the _two beams BM ₁ and BM ₂ is ₂ ° and the incident angle is symmetric, the pitch Pf of the interference fringes IF is Pf = λ / 2 sinψ as defined by the above equation (1).

On the other hand, if the pitch Pg of the diffraction grating Gw is の of that in FIG. 2, Pg = Pf. Accordingly, the diffraction angle of the ± 1st-order diffracted light (n = 1) with respect to the 0th-order light is 2
It becomes ψ. Thus, the first beam BM ₁ 0-order diffracted light B ₁
+ _1st-order diffracted light B _{1 at} every angle ± 2 ° about (0)
(+1), -1st order diffracted light B ₁ (-1), + 2nd order diffracted light B
₁ (+2), - second-order diffracted light B ₁ (-1) ......... is generated, the second beam BM ₂ 0-order diffracted light B ₂ (0) for each angle ± 2Pusai around the + 1st-order diffracted light B ₂ ( +1), -1
Order diffracted light B ₂ (-1), +2 order diffracted light B ₂ (+2), −
Second-order diffracted light B ₂ (−2) is generated. Interference fringe IF
When the pitch between the light beam and the grating Gw is equal, as shown in FIG. 10, there is no diffracted light vertically generated from the grating Gw, and the 0th-order diffracted light beam B of the beam BM ₂ traveling in the opposite direction to the beam BM _1
2 (0), the beam BM -1-order diffracted light B ₁ caused by the irradiation of ₁ (-1) and can be detected as interference light become coaxial. Similarly, the 0th-order diffracted light B ₁ (0) and the + _1st-order diffracted light B ₂ (+1) proceed coaxially, and the + _1st-order diffracted light B ₁ (+1) and the + _2nd-order diffracted light B ₂ (+2) coaxially. -1st order diffracted light B ₂ (−1) and −2nd order diffracted light B ₁
(-2) becomes coaxial. Therefore, in the device shown in FIG. 1, the 0th-order diffracted light B ₂ (0) from the grating Gw and −
The interference light BT (-1,0) with the _first-order diffracted light B ₁ (-1) is received by the photoelectric element 20b as it is, and the zero-order diffracted light B
Interference light BT between ₁ (0) and + _{1st order} diffracted light B ₂ (+1)
(0, +1) is directly received by the photoelectric element 20c.

Therefore, in the case of this embodiment, the photoelectric signal to be detected
The symbols are Sb and Sc, and the circuit block shown in FIG.
Phase difference Δθ from the reference signal SR_TwoAnd that
What is necessary is just to make the result the positional deviation amount ΔX of the grating Gw. similar
The idea is exactly the same for the transmissive reference grating 24.
The pitch of the interference fringes that apply and are created on the reference grating 24
When the pitch of the grating 24 is made equal, as shown in FIG.
Yeah, beam Br₁0th order diffracted light (directly transmitted light)
BM _TwoCoaxial with + 1st order diffracted light generated by irradiation
, The light is received by the photoelectric element 28A, and the beam B
M_Two0th order diffracted light (directly transmitted light) and beam BM₁Irradiation
And the -1st order diffracted light generated by
The light is received by the photoelectric element 28B. In this case, the reference grid 24
Since there is no diffracted light vertically generated from the beam B,
r₁, Br_TwoIs the same as that in FIG.
The distance from to the photoelectric elements 28A and 28B is as shown in FIG.
It is considerably shortened compared to, and the configuration of the reference system is compact
Can be

The two photoelectric elements 28 shown in FIG.
The output signals A and 28B may be combined by an electric analog addition (or subtraction) circuit to form a reference signal SR. Next, a fifth embodiment of the present invention will be described with reference to FIG.
FIG. 12 is a perspective view showing a configuration in which the position shift detection system of the present invention is used as an alignment system (TTR, TTL) of a step-and-repeat projection type exposure apparatus.

In FIG. 12, a reticle R is provided with a circuit pattern area PA and mark areas RMx, PMy, RMθ for TTR (through-the-reticle) alignment. The reticle R is arranged so that the optical axis AXo of the projection lens PL that is telecentric on both sides passes through the center of the pattern area PA. A dichroic mirror DCM inclined at 45 ° is disposed above the reticle R, and the exposure light is transmitted from the vertical direction to allow the pattern area PA of the reticle R to pass.
Is irradiated. The three objective lens systems GLa, GLb, and GLc of the TTR alignment system are dichroic mirrors D
Arranged to detect mark areas RMx, RMθ, and RMy of reticle R by the reflection of CM, respectively.
Here, reticle grating marks Gr having a pitch in the y direction are formed in mark regions RMθ and RMy, and reticle grating marks Gr having a pitch in the x direction are formed in mark regions RMx. Each objective lens system GLa, GLb, G
Two beams BM ₁ and BM ₂ are incident on Lc along the optical axes AXa, AXb and AXc, respectively, and cross each mark area RMx, RMθ and RMy. As shown in FIG. 13, one mark area RMx is formed as a parallelogram window in a part of the light-shielding band SB surrounding the pattern area PA, and a reticle lattice mark Gr is formed in a part of the window. ing. The parallelogram window is the reticle grating mark Gr.
This is because there is no edge extending in the y direction orthogonal to the pitch direction (x direction).

The two beams BM ₁ and BM ₂ illuminate the entire window of the mark area RMx, and the transmitted light crosses again on the wafer W via the projection lens PL. Projection lens P
The beam B is point-symmetric with respect to the optical axis AXo on the pupil plane ES of L.
The spot lights of M ₁ and BM ₂ are located. Well, wafer W
A plurality of chip areas CP are formed thereon, and a projection image of the pattern area PA of the reticle R is accurately superimposed and exposed on the chip area CP. The wafer W is placed on a stage ST that moves two-dimensionally, and mirrors IMx and IMy that reflect the laser beams of the laser interferometers LPX and LPY are fixed on two sides of the stage ST. On the other hand, the chip area CP on the wafer W has a wafer grid mark Gw
Are formed as shown in FIG. 14 corresponding to the mark areas of the reticle R. The wafer grid mark Gw is the chip area C
It is provided in a scribe line around P and is irradiated by _two beams BM ₁ and BM ₂ . In FIG. 14, RM
x ′ represents a projected image of the mark area RMx in FIG. 13, and includes a reticle lattice mark Gr and a wafer lattice mark Gw.
In FIGS. 13 and 14, y is orthogonal to the pitch direction.
The position is determined to be shifted in the direction.

In the case of such a TTR alignment system,
If the wavelengths of the _two beams BM ₁ and BM ₂ are made different from the wavelength of the exposure light to avoid absorption of the resist layer, the beams BM ₁ and BM ₂ intersect on the reticle R due to the chromatic aberration of the projection lens PL. However, it does not intersect on the wafer W (conversely, even if it intersects on the wafer W, it does not intersect on the reticle R). Therefore, a bifocal element (birefringent element) is arranged on each pupil plane (conjugate with the pupil plane ES of the projection lens PL) of the objective lens systems GLa, GLb, and GLc as in JP-A-63-283129. Reticle grating mark G
r irradiates with the P-polarized light component of the _two beams BM ₁ and BM ₂ , and the wafer grid mark Gw indicates the beams BM ₁ and BM ₂
When the irradiation is performed with the S-polarized light component, the interference light from the reticle grating mark Gr and the interference light from the wafer grating mark Gw can be separated by polarization and detected individually.

Further, as shown in FIG. 12, small mirrors Md and Me are obliquely provided immediately below the reticle R, and the objective lens system G
TTL (through-the-lens) for exclusively detecting the lattice mark Gw on the wafer W via Ld, GLe and the projection lens PL
Similarly, in the alignment system, two beams B
By irradiating the grid mark Gw with M ₁ and BM ₂ , the amount of displacement of the grid mark Gw is obtained. The TTL alignment system along the optical axis AXd detects a wafer grid mark Gw whose pitch direction is in the x direction, and the TTL alignment system along the optical axis AXe detects a wafer grid mark G whose pitch direction is in the y direction. Gw is detected.

The objective lens system G of each of these alignment systems
La, GLb, GLc, GLd, and GLe may be basically considered to be equivalent to the objective lens system 18 shown in FIG.
Although the embodiments of the present invention have been described above, the present invention can be similarly applied even when the beams BM ₁ and BM ₂ irradiating the grating Gw are of the homodyne type having no frequency difference. However, in this case, the output signals Sa, Sb, Sc are not AC signals having a constant frequency, but are sinusoidal by moving the grating Gw (wafer W) one-dimensionally (the frequency is proportional to the moving speed of the wafer W). ).

FIG. 15 shows a basic configuration for performing position shift detection or position alignment by the homodyne method.
Will be described as an example. 2 for homodyne method
There is no frequency difference between the _two beams BM ₁ and BM ₂ and the wafer W
Since the upper interference fringe is stationary, the stage ST on which the wafer W is mounted is moved one-dimensionally in the pitch direction of the grating Gw. At this time, the interference light BT (± 1) generated vertically from the grating Gw due to the relative displacement between the interference fringe and the grating Gw,
Interference light BT that follows the optical path opposite to that of beams BM ₁ and BM ₂
(-2, 0) and BT (0, +2) change in intensity. Output signals Sa, Sb, Sc of the respective photoelectric detectors are stored in a waveform storage circuit 90 for digitally sampling a signal waveform in response to an up / down pulse from a laser interferometer LPX that measures the position of the stage ST with a resolution of 0.01 μm.
a, 90b and 90c. Each storage circuit 90a, 9
Reference numerals 0b and 90c each include an analog-digital converter, a RAM, an address counter, and the like. The waveform analysis unit 92
Based on each of the stored signal waveforms, the magnitude of the amplitude is determined, the amount of displacement is detected, and the like. FIG. 16 shows the output signal Sa.
The horizontal axis represents time, and the vertical axis represents relative intensity values. The stage ST stops moving in one direction at time t ₁ , but at this time, the target positioning point A of the grid Gw
P is set to pass slightly. One cycle of the output signal Sa corresponds to a half pitch of the grating Gw here. Storage circuit 90a takes the signal waveform before the time t _1, the analyzing unit 92 detects the amplitude V _a. Similarly, the signals Sb and Sc are taken into the storage circuits 90b and 90c, and the amplitude is detected by the analysis unit 92. The analyzer 92 determines the output signal to be used in the same manner as in the first embodiment based on the amplitude value between the times t ₁ and t ₂ (stage stop period) in the same manner as in the first embodiment. The target positioning point AP based on the
_Is determined. Then, stage ST
Is returned in the reverse direction at a very low speed, and is sent to a distance within ± 1/4 pitch of the positioning point AP by the measurement value of the interferometer LPX, and at time t _S at which the slope on the waveform including the point AP appears, the stage drive motor ST is switched output signal Sa of the deviation signal level V _R to the feedback control target value.

[0061] This is the stage ST to the output signal Sa coincides with the level V _R servos, time t
Positioning after ₃ is achieved. The output signals Sb, Sc
When using, the phase of both signals will be shifted,
For example, it is necessary to perform correction based on the same concept as the method described with reference to FIGS. However, in the case of the homodyne method,
Since the information relating to the displacement amount is only the level on the output signal waveform, the level (amplitude) relationship between the two output signals Sb and Sc is corrected, and the storage circuits 90b and 9 are used.
Processing such as averaging after phase shifting (address shifting) of the waveform on at least one of the RAMs 0c is required. Such a homodyne method can be easily realized by applying the same drive frequency to the two AOMs 56 and 57 shown in FIG. In the homodyne method, a signal waveform is captured by a stage scan method, and the positional deviation amount is calculated.
There is no need to stop the stage by pre-alignment within / 4 pitch, and position detection can be performed at a relatively high speed. However, the change in intensity or fluctuation of the beams BM ₁ and BM ₂ directly causes an error. Therefore, it is necessary to measure the peak value and the bottom value of the signal amplitude immediately before the position detection.

[0062]

As described above, according to the present invention, the offset given to the measurement result can be selected, so that the measurement accuracy can be improved. In addition, by providing a plurality of modes for position measurement and making them selectable, measurement accuracy can be improved.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of a position shift detecting device according to a first embodiment of the present invention.

FIG. 2 is a diagram showing a state of generation of diffracted light for explaining the principle of the present invention.

FIG. 3 is a graph showing a diffraction angle of diffracted light in FIG.

FIG. 4 is a diagram showing a waveform of each photoelectric signal.

FIG. 5 is a view for explaining a phase shift of a photoelectric signal of three interference lights with respect to a reference signal.

FIG. 6 is a block diagram showing a specific example of a phase difference detection unit used in the first embodiment.

FIG. 7 is a diagram showing a configuration of a beam transmitting system applied to the apparatus of FIG. 1;

FIG. 8 is an optical diagram showing a modified example of the beam transmitting system applied to the apparatus of FIG. 1 and showing a second embodiment of the present invention.

FIG. 9 is a diagram showing a configuration of a position shift detecting device according to a third embodiment of the present invention.

FIG. 10 is a diagram showing a grid shape according to a fourth embodiment of the present invention.

FIG. 11 is a diagram showing a grid shape according to a fourth embodiment of the present invention.

FIG. 12 is a perspective view showing a fifth embodiment of the present invention and showing a state of a projection type exposure apparatus in which a displacement detection system is incorporated as an alignment system.

FIG. 13 is a perspective view illustrating shapes and arrangements of a reticle grating and a wafer grating used in the apparatus of FIG.

FIG. 14 is a perspective view illustrating shapes and arrangements of a reticle grating and a wafer grating used in the apparatus of FIG.

FIG. 15 is a diagram showing a configuration of a position shift detecting device according to a sixth embodiment of the present invention;

16 is a diagram showing an example of a photoelectric signal waveform obtained by the device of FIG.

[Explanation of symbols]

W ·· Wafer, Gw ·· Diffraction grating BM1, BM2 ··· Intersecting two beams, BT (± 1) ··· Interference light of ± 1st order diffracted light, BT (-2,0), BT (0, + 2) )... 0th and 2nd order diffracted light interference light, Sa, Sb, Sc... Photoelectric output signal, SR... Reference signal, 18... Objective lens system, 20a, 20b, 20c, 28. 24... Reference grating, 40, 42... Phase difference detecting section, 41, 43... Intensity detecting section, 44... Intensity judging section, 45. An acousto-optic modulator (AOM), 65 beam splitters, 70 oscillators.

Claims (4)

(57) [Claims] 1. A light beam irradiating means for irradiating a diffraction grating provided on an object to be measured with light beams from two different directions, and a photoelectric detecting means for detecting an interference intensity of specific diffracted lights generated from the diffraction grating. A detection unit, and a measurement unit that measures a displacement of the object to be measured in a grating pitch direction of the diffraction grating based on an output signal of the photoelectric detection unit. A first photoelectric detector for detecting the interference intensity of diffracted light beams having the same diffraction angle traveling in the same direction, and a second photoelectric detector for detecting the interference intensity of diffracted light beams having different diffraction angles traveling in the same direction from the diffraction grating. 2
A first measuring unit for measuring a value corresponding to the displacement of the object to be measured based on an output signal of the first photoelectric detector; and a measuring unit for measuring the value of the second photoelectric detector. A second measuring unit for measuring based on an output signal, wherein the light beam irradiating means includes a frequency modulation member for giving a predetermined frequency difference to the light beams from the two directions, and a reference to a frequency equal to the frequency difference. A reference signal generator for outputting a signal, whereby each of the first photoelectric detector and the second photoelectric detector has a first photoelectric signal and a second photoelectric signal having a frequency equal to the predetermined frequency difference. And the first measurement unit has a first phase detection unit that detects a first phase relationship between the reference signal and the first photoelectric signal as the measurement value, and the second measurement unit outputs the reference signal. A second phase relationship between the signal and the second photoelectric signal. A second phase detection unit that detects the measurement value as a measurement value, wherein the measurement unit uses the second phase difference detection unit based on a detection result of the first phase detection unit and a detection result of the second phase detection unit. A position detecting device including an offset selecting unit for giving a predetermined offset to the detection result of the position detection. 2. The offset selector according to claim 1, wherein the phase difference is 0 °,
The apparatus according to claim 1, wherein any one of + 360 ° and -360 ° is selected as the offset. 3. A diffraction grating provided on an object to be measured is irradiated with a light beam for irradiating a light beam having a predetermined frequency difference from two different directions, and a specific diffraction light generated from the diffraction grating is generated. A detection method for detecting a displacement of the object to be measured with respect to a grating pitch direction of the diffraction grating based on interference light, wherein the displacement is based on first interference light of diffraction lights having the same diffraction angle traveling in the same direction from the diffraction grating. A first phase difference between the first photoelectric signal and a reference signal having a frequency difference equal to the frequency difference, and a second interference light of diffracted lights having different diffraction angles traveling in the same direction from the diffraction grating. Detecting a second phase difference between a second photoelectric signal and the reference signal; obtaining a predetermined offset phase difference with respect to the second phase difference from the first phase difference and the second phase difference; And the office Based on the bets phase difference, the position measuring method characterized by measuring the displacement of the object to be measured. 4. An apparatus for irradiating a diffraction mark provided on an object to irradiate a light beam with a light beam, and an apparatus for measuring a displacement of the object to be measured based on optical information generated from the diffraction mark, A first photoelectric detector that detects first diffracted light information generated from the diffraction mark; a second photoelectric detector that detects second diffracted light information generated from the diffraction mark; An arithmetic unit having a plurality of modes for calculating a value corresponding to the position of the device under test using at least one of information and information from the second photoelectric detector; and selecting the plurality of modes. A position detection device comprising: a selection unit.