JPH07161611A - Position detector - Google Patents

Abstract

(57) [Abstract] [Purpose] To align the phases of the interference beat signals of light of multiple wavelengths when performing position detection using light of multiple wavelengths and using a single light receiving system. [Structure] From a laser beam from a He—Ne laser light source 1 and a laser beam from a semiconductor laser element 2,
Two pairs of luminous fluxes L1 (+ 1, -1), H1 through AOMs 8 and 9
(+ 1, -1) and L2 (-1, + 1), H2 (-1, + 1) are generated, and these two pairs of light beams are parallel flat glass 14 and relay lens 1
Irradiation is performed so as to cross the wafer mark WM via the objective lens 5 and the objective lens 17. Two pairs of diffracted light from the wafer mark WM are received by the photoelectric detector 19 via the objective lens 17 and the beam splitter 16. The rotation angle of the parallel plate glass 14 is adjusted to align the phases of the two interference beat signals.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a heterodyne interference type or homodyne interference type position detector, and more particularly to an alignment for aligning a photosensitive substrate or a mask in an exposure apparatus for manufacturing a semiconductor element, a liquid crystal display element or the like. It is suitable for application to the position detection system of an apparatus.

[0002]

2. Description of the Related Art In recent years, a pattern image of a photomask or a reticle (hereinafter collectively referred to as a "reticle") for forming a fine pattern of a semiconductor element or a liquid crystal display element on a substrate such as a semiconductor wafer or a glass plate. There is used a projection exposure apparatus that transfers a laser beam onto a substrate coated with a photoresist. In general, a semiconductor element or the like is formed by stacking a large number of circuit patterns on a substrate, and therefore a projection exposure apparatus has a circuit pattern already formed on the substrate and a reticle pattern to be exposed. ) Is provided with high accuracy. In recent years, the degree of integration of semiconductor elements such as LSI has been increasing more and more, and alignment devices are also required to perform alignment with higher accuracy. For that purpose, a position detection device that detects the positions of the reticle and the substrate with high accuracy is required.

Therefore, as an apparatus for detecting the positions of the reticle and the substrate with high accuracy, for example, Japanese Patent Laid-Open No. 2-227604.
In the publication, a heterodyne interference type position detecting device using monochromatic light is proposed. In this position detecting device, a light beam from a monochromatic light source such as a He-Ne laser light source is divided into two, and the divided two light beams are frequency-modulated by an acousto-optic modulator, respectively, so that a predetermined distance is provided between these two light beams. The frequency difference is given. In this way, the two light fluxes to which the predetermined frequency difference has been given are irradiated at a predetermined crossing angle on the diffraction grating mark as the alignment mark on the reticle. Then, by photoelectrically converting the interference light of a pair of diffracted lights generated in parallel from the diffraction grating mark, an interference beat signal on the reticle side having a frequency of the predetermined frequency difference is generated and separately detected. The position of the reticle is detected from the phase difference between the reference signal and the interference beat signal on the reticle side.

Similarly, the two light beams having such a predetermined frequency difference are irradiated onto a diffraction grating mark as an alignment mark on the substrate at a predetermined crossing angle. An interference beat signal on the substrate side is generated by photoelectrically converting the interference light of a pair of diffracted light generated in parallel from the diffraction grating mark, and the reference signal and the interference on the substrate side that are separately detected are generated. The position of the board is detected from the phase difference from the beat signal.

By the way, when the substrate is a wafer, the alignment mark is formed by an uneven pattern (phase pattern) having steps. If an attempt is made to detect such a concavo-convex pattern by the interference of monochromatic light, the diffracted light intensity becomes extremely small due to the height of the step of the alignment mark and the condition of the thin film interference of the photoresist on the wafer, and the position cannot be detected. The SN ratio of the detection signal may deteriorate.

In order to avoid such an extreme deterioration of the intensity of the diffracted light, Japanese Patent Application Nos. 5-29531 and 5-1.
In 31736, a heterodyne interference type position detection device using polychromatic light (including white light) is proposed. Among them, in the position detection device, even if white light or polychromatic light is used, it is 1
Interfering light is photoelectrically converted by one light receiving element. Further, in the position detecting device, the plurality of light receiving elements corresponding to the respective colors of the polychromatic light are used to individually photoelectrically convert the plurality of interference lights and, if necessary, the plurality of reference interference lights.

[0007]

Among the position detecting devices using polychromatic light (or white light) as described above, the position detecting device can detect the position using a single light receiving element. However, in reality, due to the manufacturing error and the mounting position error of the optical element in the light transmitting optical system for irradiating the two light fluxes at a predetermined incident angle, the photoelectric conversion of the interference light of the diffracted light of each wavelength is strictly performed. There is a disadvantage that it may be difficult to align the phases of the converted signals (interference beat signals). If the phases of the interference beat signals among a plurality of wavelengths are not aligned in such a manner, the interference beat signals obtained by photoelectric conversion with a single light receiving element (that is, the interference beat signals corresponding to the respective colors of the polychromatic light In some cases, the amplitude of (mixture) becomes small and position detection cannot be performed.

On the other hand, in the position detecting device, the interference light and the reference interference light must be photoelectrically converted for each light of each wavelength to obtain the phase of the obtained interference beat signal. There is a disadvantage that the signal processing system becomes complicated. In view of such a point, the present invention performs position detection by using a single light receiving system when performing position detection by an interference method using light of a plurality of wavelengths, and phase of the interference light of the plurality of wavelengths. It is an object of the present invention to provide a position detection device capable of aligning the positions in a predetermined state.

[0009]

As shown in FIG. 1, for example, a position detecting apparatus according to the present invention comprises a diffraction grating mark (WM) formed on a test object (18) with a plurality of coherent light beams. ) A position where the position of the object to be inspected (18) is detected based on photoelectric conversion signals of interference light of a plurality of diffracted lights generated from the diffraction grating mark (WM) by irradiating the above-mentioned plurality of light beams. In the detection device, light sources (1, 2) that generate light fluxes having different wavelengths, and a plurality of light fluxes having different wavelengths emitted from the light source are each divided into a pair of light fluxes. Optical system (5, 10, 12, 15, 17) for irradiating the diffraction grating mark (WM) formed on the object (18) with the luminous flux of
And a light receiving optical system (17, 16) for generating interference light of a plurality of pairs of diffracted light generated from a diffraction grating mark (WM) by irradiation of the plurality of pairs of light beams, and an interference light of the plurality of pairs of diffracted light Photoelectric conversion means (19) for collectively performing photoelectric conversion
And chromatic aberration correcting means (14) arranged in the light-transmitting optical system and correcting chromatic aberration of a plurality of pairs of light fluxes having different wavelengths from each other, and by the chromatic aberration correcting means (14),
The phase of the interference light of each of the plurality of pairs of diffracted light received by the photoelectric conversion means (19) is aligned in a predetermined state.

In this case, the chromatic aberration correcting means (14) may be arranged on a surface of the object (18) substantially conjugate with the surface to be inspected. Further, an example of the chromatic aberration correcting means (14) is a parallel flat plate member such as a parallel flat plate glass rotatably supported.

[0011]

According to the present invention, for example, as shown in FIG. 5, a plurality of pairs of light beams (LA1, LA2 and LB1, having different wavelengths) having different wavelengths are generated due to chromatic aberration of optical elements in the light-sending optical system.
The incident angle and the incident position of the LB2) on the diffraction grating mark (WM) are slightly changed. Therefore, if the photoelectric conversion signal of the interference light of the diffracted light by the first pair of light fluxes (LA1, LA2) is, for example, the signal (23H) of FIG. 2B, the second pair of light fluxes (LB1, LB2) The photoelectric conversion signal of the interference light of the diffracted light is, for example, the signal (23
L) and the phases are different from each other, so that the photoelectric conversion signal obtained by receiving light collectively by the photoelectric conversion means (19) has a small amplitude like the signal (23) in FIG. 2D. , The contrast decreases and the SN ratio deteriorates.

Therefore, by the chromatic aberration correcting means (14),
By correcting the state of chromatic aberration, the first pair of light fluxes (LA1, LA
The photoelectric conversion signal of the interference light of the diffracted light by 2) is shown in FIG.
As the signal (24H) of (b), the second pair of luminous flux (LB
1 and LB2), the photoelectric conversion signal of the interference light of the diffracted light is set as the signal (24L) in FIG. As a result, the photoelectric conversion signal obtained by receiving light collectively by the photoelectric conversion means (19) has a large amplitude like the signal (24) in FIG. 2D, and the position detection is performed with high accuracy.

Further, when the chromatic aberration correcting means (14) is arranged on a surface which is substantially conjugate with the surface to be inspected of the object (18), the chromatic aberration correcting means (14) can be made small and the detection device as a whole. Can be downsized. Next, chromatic aberration correction means (1
As 4), the operation of the parallel flat plate glass when the parallel flat plate glass supported rotatably is used will be described. In this parallel plate glass, light of each wavelength is refracted according to Snell's law, and the position of each light beam shifts when it is emitted from the parallel plate glass. The shift amount Δ ₁ of the light flux of the first wavelength λ ₁ is expressed as follows when the parallel plate glass has a thickness t and the refractive index n _{1 and} the incident angle of the light flux is φ.

Δ₁ = Tsin φ {1-cos φ / (n₁cos φ₁′)} (1) where sin φ = n₁sin φ₁′ (2) Further, the wavelength λ₂Amount of light flux shift Δ₂Also parallel flat
The refractive index of the plate glass is n ₂Can be expressed as
Wear. Δ₂ = Tsin φ {1-cos φ / (n₂cos φ₂′)) (3) where sin φ = n₂sin φ₂′ (4) Therefore, the wavelength λ₁, And λ₂Relative shift amount of light flux (Δ
₁-Δ₂) Is as follows.

Δ₁-Δ₂= Tsin φcos φ {(1 / (n₂cos φ₂') -1 / (n₁cos φ₁′)} (5) This is determined according to the incident angle φ with respect to the parallel plate glass.
Diffraction angle φ at each wavelength ₁'And φ₂'By two waves
Change in relative shift amount between long light beams, that is, parallel
Light of two wavelengths can be adjusted by adjusting the rotation angle of flat glass.
This means that the relative phase of the electrical conversion signal can be adjusted. this
Allows easy duplication with a simple configuration and simple adjustment.
It is possible to align the phases of photoelectric conversion signals of light fluxes of several wavelengths.
Wear.

[0016]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A first embodiment of the position detecting device according to the present invention will be described below with reference to FIGS. In this embodiment, the present invention is applied to an off-axis type alignment system provided independently of a projection optical system for exposure. FIG. 1 shows the position detecting device of this embodiment.
2, a laser beam H having a wavelength λ ₁ emitted from the He—Ne laser light source 1 and a laser beam L having a wavelength λ ₂ emitted from the semiconductor laser element 2 and collimated by the collimator lens 3 are supplied to the dichroic mirror 4. Are coaxially combined and irradiated to the diffraction grating 5.

± from the diffraction grating 5 by the laser beam H
Diffraction by first-order diffracted lights H1 and H2 and laser beam L
Spatial fill of ± 1st order diffracted lights L1 and L2 from the grating 5
Relayed the two pairs of extracted laser beams
An acousto-optic modulator (hereinafter referred to as “AO
It irradiates so that it may cross in 8). Frequency f
₁In AOM8 to which the drive signal of
Traveling to the right, this ultrasonic wave diffracts the laser beam.
When performing (acoustic Bragg diffraction), each is caused by the Doppler effect.
The laser beam is frequency modulated. At this time, +1 order
+ 1st-order diffracted light L1 by the AOM8 of the diffracted lights L1 and H1
(+1) and H1 (+1) are in the form of going to the ultrasonic wave
Therefore, the frequency becomes higher. However, the speed of light is supersonic
It is sufficiently larger than a wave, and the direction of travel of ultrasonic waves and the direction of travel of light
The angle with the direction is almost right, so it is ejected from AOM8
The frequency of the + 1st order diffracted light is
Frequency f₁It will be higher by the amount.

On the other hand, the AOM of the -1st order diffracted lights L2 and H2
The -1st-order diffracted lights L2 (-1) and H2 (-1) by 8 have a shape in which they travel in the same direction as the ultrasonic waves, so that the frequency becomes low and AO
The frequency of the −1st order diffracted light emitted from M8 is AOM8.
The frequency becomes lower by the frequency f ₁ of the drive signal of. Also, AOM
Since diffracted light other than ± 1st order diffracted light is generated within 8,
Diffracted lights L1 and H incident on the traveling direction of ultrasonic waves in M8
A so that the incident planes of 1, L2 and H2 intersect at 45 °
The OM8 is rotated and arranged, and a spatial filter (not shown) is arranged in the vicinity of the relay lens 7B immediately after the AOM8. As a result, the + 1st-order diffracted light L1 (+1), H1 (+1) by the AOM8 of the + 1st-order diffracted light L1, H1 and the -1st-order diffracted light L2 (-1) by the AOM8 of the -1st-order diffracted light L2, H2,
Only H2 (-1) is taken out.

+ 1st order diffracted lights L1 (+1), H1 (+1) and -1st order diffracted lights L2 (-1), which are emitted from the relay lens 7B.
H2 (-1) enters so as to cross into the AOM 9 to which the drive signal of frequency f ₂ (≠ f ₁ ) is applied. In AOM9, the ultrasonic wave is traveling from right to left, and the + 1st order diffracted light L1 (+
1) and H1 (+1) by the AOM9 -1st order diffracted light L1 (+1,
The frequencies of -1) and H1 (+ 1, -1) are lowered by the frequency f ₂ of the drive signal of the AOM 9. On the other hand, the -1st order diffracted light L2
+ 1st-order diffracted light L2 (-by AOM9 of (-1) and H2 (-1)
The frequencies of 1, + 1) and H2 (-1, + 1) are increased by the frequency f ₂ of the drive signal of the AOM 9.

Further, in the AOM 9 other than ± first-order diffracted light
Since diffracted light is generated, the direction of travel of ultrasonic waves in the AOM 9
A so that the incident surface of the incident diffracted light intersects at 45 °
OM9 is rotated and placed, and the relay immediately after AOM9
-Place a spatial filter (not shown) near the lens 10.
It As a result, the −1st order diffracted light (hereinafter, simply referred to as “light flux”)
L1 (+ 1, -1), H1 (+ 1, -1), and + 1st order diffracted light
(Hereinafter, simply referred to as "light flux") L2 (-1, + 1), H2 (-1,
Only +1) is fetched. In this case, the wavelength λ₁The laser
The frequency of beam H is f_Ten, Wavelength λ₂Of the laser beam L
Frequency is f₂₀Then, the luminous fluxes L1 (+ 1, -1) and H1 (+1,
The frequencies of -1) are (f₂₀+ F₁-f ₂) And (f_Ten+
f₁-f₂), And the luminous flux L2 (-1, + 1), H2 (-1, + 1)
The wave number is (f₂₀-F₁+ f₂) And (f_Ten-F₁+ f₂)
Becomes Therefore, the luminous flux L2 (-1, + 1) and the luminous flux L1 (+ 1, -1)
Between the beat frequency and the luminous flux H2 (-1, + 1) and luminous flux H
The beat frequency between 1 (+ 1, -1) is 2 | f₁-f
₂|, That is, the difference in drive frequency between the two AOMs 8 and 9
Doubled. Regarding the detailed derivation process, Japanese Patent Application No.
No. 24441 is also disclosed.

Generally, the drive frequency of the AOM is on the order of MHz, while the beat frequency that can be processed by the two-beam interference is on the order of kHz. So two AOMs
As the driving frequencies f ₁ and f ₂ of 8 and 9, for example, a frequency of several tens MHz and a frequency difference of several tens kHz are used, respectively, a beat frequency that can be easily processed can be obtained.

Wavelength λ ₁ emitted from the relay lens 10
Pair of luminous fluxes H2 (-1, + 1), H1 (+ 1, -1) and wavelength λ ₂
Pair of light fluxes L2 (-1, + 1) and L1 (+ 1, -1) enter the beam splitter 11, and two pairs of diffracted light transmitted through the beam splitter 11 are transmitted by the relay lens 12 to the field stop 13
The light is focused so that it intersects on the opening. Then, the light flux that has passed through the aperture of the field stop 13 has a parallel plate glass 14,
A collimator lens 15, a beam splitter 16 for separating the transmitted light and the received light, and an alignment mark in the form of a diffraction grating (hereinafter referred to as a "wafer mark") on a wafer 18 which is a position detection target via an objective lens 17. ) WM
It is incident so that it crosses over. That is, the arrangement surface of the field stop 13 is conjugate with the surface of the wafer 18, the image of the aperture of the field stop 13 is relayed to the surface of the wafer 18, and the parallel flat glass 14 is near the surface conjugate with the surface of the wafer 18. It is arranged.

[0023] In this case, the pitch direction in the X direction of the wafer mark WM, the direction parallel to the optical axis of the relay lens 15 and 17 and the Z direction, the light beam of a pair of wavelength lambda ₁ H1 (+ 1, -
1) and H2 (-1, + 1) are diffraction grating wafer marks W
Irradiation is performed from two directions so as to intersect with M along the X direction, and thus interference fringes that flow along the pitch direction (X direction) are generated on the wafer mark WM. And
In the normal direction (Z direction) of the diffraction grating mark WM, the + 1st order diffracted light H2 (-1, + 1, + 1) and the light flux H1 (+ 1, -1) of the light flux H2 (-1, + 1). -1st-order diffracted light H1 (+ 1, -1, -1) is generated in parallel.

Here, the luminous fluxes H1 (+ 1, -1) and H2 (-1, +)
1) illuminates the wafer mark WM from two directions, the crossing angle is P _WM , the pitch of the wafer mark _WM , and the luminous flux H1 (+
When the incident angle of 1, -1) or H2 (-1, + 1) with respect to the wafer mark WM is θ _WM1 , it is set so as to satisfy the following condition. _{sin θ WM1 = λ 1 / P} WM (6) Similarly, the light beam L1 (+ 1, -1) of a pair of wavelength lambda ₂ and L2 (-
(1, + 1) is also irradiated from two directions so as to intersect the diffraction grating wafer mark WM along the X direction, and the light beam L2 (in the normal direction (Z direction) of the diffraction grating mark WM). -1,
+1) + 1st order diffracted light L2 (-1, + 1, + 1) and luminous flux L1 (+ 1,-
The -1st-order diffracted light L1 (+ 1, -1, -1) of 1) is generated in parallel.
The crossing angle when the light fluxes L1 (+ 1, -1) and L2 (-1, + 1) illuminate the wafer mark WM from two directions is the light flux L
When the incident angle of 1 (+ 1, -1) or L2 (-1, + 1) with respect to the wafer mark WM is set to θ _WM2 , the following conditions are set to be satisfied.

Sin θ _WM2 = λ ₂ / P _WM (7) As a result, two pairs of diffracted light (H2 (-1, + 1, + 1), H1 (+ 1, -1,) generated from the wafer mark WM are generated. -1) and L2 (-1, + 1, +
1) and L1 (+ 1, -1, -1)) are reflected again by the beam splitter 16 after passing through the objective lens 17, and then the photoelectric detector 19
Is converted into a wafer signal SW having a frequency of 2 | f ₁ -f ₂ |.

On the other hand, the two pairs of light beams reflected by the beam splitter 11 intersect on the reference grating 21 by the condenser lens 20 of the reference optical system for determining the origin of the phase of the beat signal. The pitch of the reference grating 21 is determined so that the ± 1st-order diffracted light of each light flux proceeds in the normal direction (vertical direction), and the interference light composed of the two pairs of ± 1st-order diffracted light is photoelectrically converted by the reference photoelectric detector 22. And the frequency is 2 | f ₁ -f ₂ |
The reference signal SR of is obtained. From the phase shift of the wafer signal SW based on the reference signal SR, X of the wafer 18 is detected.
The position in the direction is calculated by a signal processing system (not shown).

Next, the operation of the parallel plate glass 14 of FIG. 1 will be described. FIG. 5 shows a case in which the parallel plate glass 14 is not provided in FIG. 1 and the beam splitter 16 for separating light and light is incomplete. In FIG. 5, a beam splitter 16A for transmitting / receiving light is an example of a beam splitter whose incident surface and outgoing surface are not completely parallel. In this case, the wafer marks of the light beams LB1 and LB2 of wavelength λ ₁ are used. Crossing position on WM and light flux L of wavelength λ ₂
The intersection positions of A1 and LA2 on the wafer mark WM are different.

[0028] Then, when the parallelism between the incident surface and an exit surface of the beam splitter 16A is different only α at the apex angle of the triangular prism, a beam splitter for the wavelength lambda ₁ 1
Assuming that the refractive index of 6A is n ₁ , the refractive index of the beam splitter 16A for the wavelength λ ₂ is n ₂ , and the focal length of the objective lens 17 is F, the crossing position of the light flux of the wavelength λ ₁ on the surface of the wafer 18 The displacement amount δ ₁ is as follows. δ ₁ = F (n ₁ −1) α (8) Similarly, the positional shift amount δ ₂ of the crossing position of the light flux of wavelength λ ₂ on the surface of the wafer 18 is as follows.

Δ ₂ = F (n ₂ −1) α (9) Therefore, for the light flux of wavelength λ _{1 and} the light flux of wavelength λ ₂ , the intersection position on the wafer 18 is relatively expressed in the following formula in the X direction. It deviates by just such a value. δ ₁ −δ ₂ = F (n ₁ −n ₂ ) α (10) This is indicated by the wafer signal SW output from the photoelectric detector 19 and the reference signal SR output from the reference photoelectric detector 22 in FIG. 1. Then, while the reference signal SR becomes the periodic signal shown in FIG. 2A, the wafer signal SW becomes as shown in FIG.
The signal is as shown by the waveform 23 of the solid line. Then, of the wafer signal SW, the interference beat signal SW _H of the wavelength λ ₁
2B has a waveform 23H indicated by a solid line in FIG. 2B, and the interference beat signal SW _L having the wavelength λ ₂ has a waveform 23L indicated by a solid line in FIG. 2C, which interferes with the interference beat signal SW _H. The phase of the beat signal SW _L differs from that of the beat signal SW _{L by an amount corresponding} to the relative positional deviation of the equation (10). Therefore, the positional deviation amount of the detected signal varies depending on the wavelength, and the positional deviation amount changes depending on the interference condition of the resist film applied on the wafer, which makes accurate alignment difficult.

Therefore, in order to detect the interference beat signal of the same phase for all wavelengths by the photoelectric detector 19 of FIG. 1, the positional deviation of the beam intersections due to the manufacturing error of the optical parts as shown in FIG. It has to be corrected. Therefore, in this embodiment, as shown in FIG.
Surface (image space) near the surface in the light-transmitting optical system that is conjugate with the surface of the
A parallel flat plate glass 14 made of glass having color dispersion (or optical plastics or the like is also possible) is arranged, and the parallel flat plate glass 14 is the optical axis direction (Z direction) of the light sending optical system and the X direction. It is rotatably supported about a shaft 14a perpendicular to the direction. Then, the rotation angle of the parallel plate glass 14 around the axis 14a is adjusted through an adjusting device (not shown), so that the displacement of the intersecting position of the two pairs of light beams on the wafer mark WM is corrected. to correct.

Here, an example of a method for adjusting the parallel flat plate glass 14 in the position detecting device shown in FIG. 1 will be described.
First, in FIG. 1, the He-Ne laser light source 1 and the semiconductor laser element 2 are turned on, the driving of the AOMs 8 and 9 is started, and the He-Ne laser light source 1 and the semiconductor laser element 2 are stabilized with sufficient time. Make it oscillate. Next, a reference diffraction grating having a wide area serving as a reference having a pitch equal to the wafer mark WM is arranged on the surface of the wafer 18. The reference diffraction grating for this adjustment may be an intensity grating (amplitude grating) manufactured by etching a deposited film of chromium or the like,
The size may be enough to cover the shift amount of the light flux due to the parallel plate glass 14 for correcting chromatic aberration. He-Ne now
Two pairs of (2
When a light beam (of a wavelength) (heterodyne beam) is applied to the reference grating 21 and the standard diffraction grating on the wafer 18, the wafer signal SW (interference beat) from the photoelectric detector 19 shown by the waveform 23 in FIG. signal), to measure the phase difference phi _all with respect to the reference signal SR from the reference photodetector 22 shown in FIG. 2 (a).

Next, in front of the photoelectric detector 19 of FIG.
Search for a filter that transmits only one of _{1 and} λ ₂ (for example, a colored glass filter that transmits only the laser beam of wavelength λ ₂ from the semiconductor laser element 2), and ) Beat signal SW _L shown by waveform 23L
The measured phase difference phi _L with respect to the reference signal SR, the former so that the phase difference phi _all latter phase difference phi _L equal, rotate the parallel plate glass 14. When the light emission amounts of the He-Ne laser light source 1 and the semiconductor laser device 2 are equal, the phase difference φ _all measured using both lights is shown in FIG. 2B measured using the light of wavelength λ ₁ . It is considered to be the average of the phase difference φ _H shown and the phase difference φ _L measured using the light of wavelength λ ₂ .

At this time, the reference signal SR must always be photoelectrically converted in a state where the interference beat lights of both wavelengths are mixed. That is, only one of the He—Ne laser light source 1 and the semiconductor laser element 2 should not be turned on to measure the phase difference of signals at each wavelength. Also, if it is not possible to measure only the phase difference of the light signal of one wavelength due to the characteristics of the filter (color glass filter), the phase difference φ
Assuming _all as follows, the rotation amount θ of the parallel plate glass 14 is obtained.

Φ _all = (φ _L + φ _H ) / 2 (11) The rotation amount θ of the parallel flat plate glass 14 is t when the thickness of the parallel flat plate glass 14 is from the semiconductor laser element 2 on the parallel flat plate glass 14. The refractive index for light of wavelength λ ₂ is n _L , and the wavelength λ ₁ from the He-Ne laser light source 1 on the parallel plate glass 14 is
The refractive index of the light at n _H and the pitch of the wafer mark WM at P
_WM , the magnification from the field stop 13 of the light-transmitting optical system to the surface of the wafer 18 is β, and is as follows.

Βt sin θ (1 / n _L −1 / n _H ) ≈ {P _WM / (4π)} (φ _H −φ _L ) ≈ {P _WM / (2π)} (φ _all −φ _L ) (12 Actually, since the parallel plate glass 14 also has a manufacturing error, after rotating by the calculated rotation amount θ, the phase difference when the light from both light sources is used again, and Measure the phase difference when using the light from one light source, and confirm that they match. If they do not match, the rotation amount of the parallel flat plate glass 14 is calculated again, and
Repeat the work of turning that amount. Accordingly, when the reference diffraction grating on the wafer 18 is at the reference position,
The interference beat signal SW _{H having} the wavelength λ ₁ and the interference beat signal SW _L having the wavelength λ _{2 have} a waveform 24 H shown by a dotted line in FIG. 2B and a waveform 24 L shown by a dotted line in FIG. The phase difference becomes zero. Therefore, the wafer signal SW
2 becomes a waveform 24 shown by a dotted line in FIG. 2D, that is, a signal having a large amplitude and a good SN ratio, and position detection is performed with high accuracy.

Thereafter, as the wafer 18, a wafer mark WM whose positional relationship with the position of the pattern formed in the previous steps (that is, the position where the pattern of the reticle is transferred next) is accurately known is formed. 1 is used and the position of the wafer stage on which the wafer 18 is mounted is measured in the X direction by an interferometer (not shown). It is detected by the device.

Then, the measured value of the interferometer when the shot area to which the wafer mark WM belongs is set in the exposure field of the projection optical system (not shown), and the position of the wafer mark WM is previously detected. A value obtained by adding the positional deviation amount of the wafer mark WM measured by the position detection device to the difference from the measurement value of the interferometer is stored as an offset (baseline). In the actual exposure, after the position detecting apparatus measures the positional deviation amount of the wafer mark WM, the wafer 18 is moved by an amount corresponding to the measured positional deviation amount plus the previous offset, and then the exposure is performed.

As described above, according to this embodiment, since the phases of the interference beat signals of two wavelengths are aligned by the parallel flat glass 14, the contrast of the wafer signal SW which is a mixture of the interference beat signals of two wavelengths. Becomes higher, and position detection is always performed with high accuracy. Next, a second embodiment of the present invention will be described with reference to FIGS. In this example,
The present invention is applied to a TTR (through the reticle) type alignment system, and in FIGS. 3 and 4, the portions corresponding to those in FIG. 1 are denoted by the same reference numerals and detailed description thereof will be omitted.

FIG. 3 (a) is a front view of the heterodyne beam generation system of this embodiment, and FIG. 3 (b) is a side view of FIG. 3 (a). In FIGS. 3 (a) and 3 (b), He-N
e A laser beam H of wavelength λ ₁ emitted from the laser light source 1 and a laser beam L of wavelength λ ₂ emitted from the semiconductor laser element 2 and collimated by the collimator lens 3 are coaxially combined by the dichroic mirror 4A. Then, the diffraction grating 5 is irradiated.

± 1st-order diffracted lights H1 and H from the diffraction grating 5
2 and the ± 1st-order diffracted lights L1 and L2 to the relay lens 3
1A, through another spatial filter 32 for blocking diffracted light and a relay lens 31B, irradiation is performed so as to intersect the AOM 8 driven at the frequency f ₁ . Since diffracted lights other than the ± 1st-order diffracted lights are also generated in the AOM 8, diffracted lights L1, H1, and L that are incident along the traveling direction of the ultrasonic waves in the AOM 8 are incident.
AOM8 so that the plane of incidence of 2 and H2 intersect at 45 °
Is rotated and placed, and the spatial filter 34 is placed near the relay lens 33A immediately after the AOM 8. As a result, the + 1st-order diffracted light L1 (+1), H1 (+1) and the -1st-order diffracted light L2, H due to the AOM8 of the + 1st-order diffracted light L1, H1.
Only the -1st order diffracted lights L2 (-1) and H2 (-1) by the AOM 8 of 2 are extracted.

Relay lens 33 following spatial filter 34
The + 1st-order diffracted lights L1 (+1) and H1 (+1) and the -1st-order diffracted lights L2 (-1) and H2 (-1) emitted from B have frequencies f ₂ (≠ f
It enters so as to cross into the AOM 9 to which the drive signal of ₁ ) is applied. Since diffracted lights other than the ± 1st-order diffracted lights are generated also in the AOM 9, the AOM 9 is rotated and arranged so that the traveling direction of the ultrasonic wave in the AOM 9 and the incident surface of the incident diffracted light intersect at 45 °, and the AOM 9 The spatial filter 36 is arranged near the relay lens 35A immediately after. The spatial filter 36 allows the -1st-order diffracted light (hereinafter, simply referred to as "light flux") L1 (+ 1, -1), H1 (+ 1, -1), and the + 1st-order diffracted light (hereinafter, simply referred to as "flux") L2. (-1, + 1), H2 (-1,
Only +1) is fetched. In this case, the luminous flux L2 (-1, + 1)
Frequency between the light flux L1 (+ 1, -1) and the light flux H
The beat frequencies between 2 (-1, + 1) and the luminous flux H1 (+ 1, -1) are both 2 | f ₁ -f ₂ |, that is, the difference of the drive frequencies between the two AOMs 8 and 9. Doubled.

Structures following FIGS. 3 (a) and 3 (b) are shown in FIGS. 4 (a) and 4 (b). In FIG.
Luminous flux L1 (+ 1, -1), L2 (-1, + 1), H1 (+ 1, -1) in FIG.
And H2 (-1, + 1) are represented by luminous fluxes LA, LB, HA and HB, respectively. In addition, FIG.
Let (b) be the ZX plane. These FIGS. 4 (a) and 4 (b)
, Two pairs of light beams (LA, LB and HA, H emitted from the relay lens 35B so as to intersect each other).
B) shows two pairs of light fluxes for the reticle (LAR, LBR and HAR, HBR) and two pairs of light fluxes for the wafer (LAW, LBW and HAW, HBW) by the beam splitter 37.
Is divided into and

Then, two pairs of light beams (LAW,
LBW and HAW, HBW) are parallel plate glass 38,
After passing through the beam splitter 41 after being formed into two pairs of parallel light fluxes through the relay lens 40, the objective lens 42 causes the objective lens 42 to form a diffraction grating-shaped alignment mark (reticle mark) RM on a window portion near the RM. Collected. After that, the two pairs of light beams for the wafer are projected onto the projection optical system 44.
Thus, the wafer marks WM on the wafer 18 are irradiated at predetermined crossing angles. Then, two pairs of diffracted light (first-order diffracted light LAW (+1) of the light beam LAW, −1st-order diffracted light LBW (-1) of the light beam LBW, and the light beam are emitted from the wafer mark WM in the direction (normally) of the normal line. HAW + 1st order diffracted light HAW (+1), luminous flux HBW -1st order diffracted light HBW (-1))
Returns to the window of the reticle 43 via the projection optical system 44.

On the other hand, two pairs of light fluxes for the reticle (LAR,
LBR and HAR, HBR) are parallel plate glass 39,
After passing through the beam splitter 41 after being formed into two pairs of parallel light fluxes via the relay lens 40, the objective lens 42 irradiates the reticle mark RM in the form of a diffraction grating on the reticle 43 with a predetermined crossing angle. It Then, from the reticle mark RM almost in the normal direction (vertically)
Two pairs of diffracted light emitted (-first-order diffracted light L of light flux LAR
AR (-1), + 1st order diffracted light LBR (+1) of the light beam LBR, -1st order diffracted light HAR (-1) of the light beam HAR, +1 of the light beam HBR
The second-order diffracted light HBR (+1) is two pairs of diffracted light (LAW (+1), LBW (-1) and HAW (+1), H from the wafer mark WM.
After being reflected by the beam splitter 41 through the objective lens 42 in parallel with BW (-1)), the light enters the beam splitter 46 (or the partial reflection prism) through the condenser lens 45.

The interference light of the two pairs of diffracted light from the wafer mark WM reflected by the beam splitter 46 is received by the wafer photoelectric detector 47, and the wafer photoelectric detector 4 is detected.
The wafer signal SB obtained by the photoelectric conversion in 7 is supplied to the control device 50. At the same time, the interference light of two pairs of diffracted light from the reticle mark RM that has passed through the beam splitter 46 is received by the reticle photoelectric detector 49 and is obtained by photoelectric conversion in the reticle photoelectric detector 49. The signal SA is supplied to the control device 50.

The reticle 43 is the reticle stage R.
The wafer 18 is placed on the ST and the wafer 18 is placed on the wafer stage WST.
Mounted on the reticle stage RST, the reticle 43 performs translation and rotation adjustment of the reticle 43 within a predetermined range, and the wafer stage WST performs stepping drive of the wafer 18 in the X and Y directions and position adjustment in the Z direction. . Control device 5
0 is the supplied reticle signal SA and wafer signal SB
Based on the above, the reticle 43 and the wafer 18 are aligned with each other via the reticle stage RST and the wafer stage WST.

Also in this embodiment, since the photoelectric detector 47 for wafer photoelectrically converts the interference beat lights of two wavelengths in a mixed state, the parallel plate glass 38 is used to match the phases of the two interference beat lights. To rotate. Also for the reticle photoelectric detector 49, the parallel plate glass 39 is rotated to match the phases of the two interference beat lights. Since the wafer alignment mark may be destroyed during the semiconductor process, it may be necessary to replace the mark. for that purpose,
It is necessary to move the alignment system including the objective lens 42 or a part of the alignment system parallel to the reticle surface so that the rewritten mark can be detected. In this case,
The alignment system or a part thereof can be moved so as not to change the relative positional relationship between the reticle beam and the reticle mark. With this movement, the image height when the wafer beam passes through the projection lens 44 (projection optical System 44
The height of the reticle beam from the projection optical system 44 changes as a result. Therefore, the parallel flat glass 38 may be rotated to adjust the wafer beam in order to position the origin (reference point) of the optical interference beat signal obtained through the wafer mark each time the wafer mark is changed. However, it is sufficient to adjust the reticle beam only once. When the alignment system is based on the homodyne method and the wafer beam is adjusted each time the wafer mark is changed, the parallel plate glass 38 is rotated so as to reach the peak position of the detection signal obtained through the wafer mark. Then, the wafer beam may be adjusted.

Next, the parallel plate glass 3 in this embodiment is used.
The adjustment method of No. 8 will be described. First, a reference diffraction grating is set on the wafer 18, and the position of the reticle mark RM is aligned with that of the reference diffraction grating by observing with an alignment microscope (not shown). In this state, the alignment microscope is retracted so that the reticle mark RM and its reference diffraction grating are irradiated with two pairs of light fluxes for the reticle and two pairs of light flux for the wafer, respectively. After that, for example, only the semiconductor laser element 2 of FIG. 3 is turned on, and 1 is turned on the reticle mark RM and its reference diffraction grating.
The pair of light beams LAR, LBR and LAW, LBW are irradiated. At this time, the reticle photoelectric detector 49 and the wafer photoelectric detector 47 obtain a reticle signal SA _L and a wafer signal SB _L having a frequency of 2 | f ₁ -f ₂ |, respectively, and the reticle 43 of the reference diffraction grating. The amount of deviation in the X direction with respect to
It is measured as the phase difference between wafer signal SB _{L and} reticle signal SA _L.

In order to actually adjust the rotation angle of the parallel plate glass 38, the semiconductor laser element 2 of FIG. 3 is turned on,
The phase difference Δ _ABL of the wafer signal SB _L with respect to the reticle signal SA _{L in} a state where the He-Ne laser light source 1 is turned off.
After measuring the reticle, the reticle signal S from the reticle photoelectric detector 49 is also turned on with the He-Ne laser light source 1 also turned on.
Wafer signal S from wafer photoelectric detector 47 for A
Measure the phase difference Δ _{AB of} B. Then, by rotating the parallel plate glass 38 such that the phase difference delta _ABL and the phase difference delta _AB equal to the phase difference between the reticle signal SA and the wafer signal SB at the time and offset.

After that, when actually performing the alignment,
He-Ne laser light source 1 and semiconductor laser device 2 of FIG.
With lit, reticle stage RST or wafer stage WST is moved so that the phase difference between reticle signal SA and wafer signal SB becomes the offset.
When performing the alignment by the TTR method as described above, the wafer mark WM is likely to be exposed by the exposure light, so that the wafer mark WM needs to be replaced. In this case, the distance (offset) between the adjacent wafer marks is measured after the field of view of the position detection device is moved, and the position shifted by the distance is set in the exposure field.
Exposure may be performed after moving wafer stage WST or reticle stage RST.

Further, in order to operate the laser light source stably, instead of turning off the He--Ne laser light source 1 when adjusting the parallel plate glass 38, a He or the like is used by using a shutter or the like.
-Shield the laser beam from the Ne laser light source 1 or
Alternatively, the measurement may be performed by dimming the laser beam from the He—Ne laser light source 1 with an ND filter.
Instead of opening and closing the light beam from the He-Ne laser light source 1 with a mechanical shutter, an optical element such as a Pockels cell may be used to form the shutter.

As a light source, a He-Ne laser light source 1
Instead of using such a laser light source and the semiconductor laser element 2 in combination, one laser light source that simultaneously oscillates a plurality of wavelengths may be used as a light source. Further, as the light source, a light source (white lamp, laser diode) or the like that generates light with a wide spectrum width may be used. Further, the present invention is not limited to an off-axis type alignment system or a TTR type alignment system, but also a TT for performing position detection via a projection optical system.
The same applies to the L (through the lens) type alignment system.

Further, although the present invention is applied to the heterodyne interference type alignment system in the above-mentioned embodiment, the present invention can be applied to the homodyne interference type alignment system as well. Thus, the present invention is not limited to the above embodiment,
Various configurations can be adopted without departing from the scope of the present invention.

[0054]

According to the present invention, even if a single light receiving system is used while simultaneously using lights having a plurality of wavelengths, the phase of the interference lights having a plurality of wavelengths can be set to a predetermined state by the chromatic aberration correcting means. They can be aligned, and the SN ratio of the detection signal from the single light receiving system can be improved. Therefore, the position can be detected with high accuracy without being affected by the step of the diffraction grating mark or the film thickness of the photosensitive material, and even if an optical element including a manufacturing error is used, the phase shift caused by the manufacturing error is caused. There is an advantage that can be corrected.

Further, when the chromatic aberration correcting means is arranged on a surface which is substantially conjugate with the surface to be inspected of the object to be inspected 18, the entire apparatus can be downsized. Further, when the chromatic aberration correcting means is a parallel flat plate glass which is rotatably supported, the structure is simple.

[Brief description of drawings]

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

FIG. 2 is a waveform diagram showing a reference signal and a wafer signal obtained in the embodiment of FIG.

FIG. 3A is a front view of a heterodyne beam generation system according to a second embodiment of the present invention, and FIG. 3B is a side view of the heterodyne beam generation system according to the second embodiment.

FIG. 4 (a) is a front view showing the configuration of a portion following FIG. 3 (a) of the second embodiment of the present invention, and FIG. 4 (b) is a portion of FIG. 3 (b) of the second embodiment thereof. It is a side view which shows a structure.

FIG. 5 is an explanatory diagram of a positional deviation amount at an intersection of two light fluxes when an optical element including a manufacturing error is used.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 He-Ne laser light source 2 Semiconductor laser element 3 Collimator lens 4,4A Dichroic mirror 5 Diffraction grating 8,9 Acousto-optic modulator (AOM) 13 Field stop 14 Parallel plate glass 18 Wafer WM Wafer mark 19 Photoelectric detector 21 Reference grating 22 Reference Photoelectric Detector 43 Reticle RM Reticle Mark 47 Photoelectric Detector for Wafer

Claims (3)

[Claims] 1. A plurality of coherent light beams are applied to a diffraction grating mark formed on a test object, and a plurality of diffracted lights generated from the diffraction grating mark by the irradiation of the plurality of light beams. In a position detecting device for detecting the position of the object to be detected based on the photoelectric conversion signal of the interference light, a light source that generates light fluxes having a plurality of different wavelengths, and a plurality of different wavelengths emitted from the light source. A light-sending optical system that splits each of the light fluxes into a pair of light fluxes, and irradiates a plurality of pairs of light fluxes obtained thereby onto a diffraction grating mark formed on the object to be inspected; A light receiving optical system that generates interference light of a plurality of pairs of diffracted light generated from the diffraction grating mark by irradiation, a photoelectric conversion unit that collectively photoelectrically converts the interference light of the plurality of pairs of diffracted light, and the light transmission Located in the optics, front Chromatic aberration correcting means for correcting chromatic aberration of a plurality of pairs of light fluxes having different wavelengths, and by the chromatic aberration correcting means, the phase of each interference light of the plurality of pairs of diffracted light received by the photoelectric conversion means is determined. A position detecting device, which is arranged in a predetermined state. 2. The position detecting device according to claim 1, wherein the chromatic aberration correcting means is arranged on a surface substantially conjugate with a surface to be inspected of the object to be inspected. 3. The chromatic aberration correcting means is a parallel flat plate glass which is rotatably supported.
Alternatively, the position detection device according to item 2.