JP3031321B2 - Projection exposure apparatus and 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 projection exposure apparatus used in a lithography process of a semiconductor device, a liquid crystal display device, and the like, and more particularly, to an alignment error caused by a distortion (distortion error) generated in a projection system of the projection exposure device. The present invention relates to an alignment apparatus in which is reduced.

[0002]

2. Description of the Related Art In recent years, a stepper equipped with a projection lens having a high numerical aperture has been used as an apparatus for printing a pattern of a mask (reticle) on a semiconductor wafer with a submicron resolution. In this type of stepper, a chip pattern (shot area) already formed on a wafer and a reticle pattern to be newly superposed and exposed must be aligned with a total accuracy of a fraction of the minimum line width or less. For this reason, in recent years, a stepper equipped with a more accurate alignment device (sensor) has been studied, and development for practical use has been promoted.

[0003] In the future, 4Mbit DRAM, 16Mbit
A stepper, which is the mainstream in the production of DRAMs, sequentially detects marks on a reticle and marks in each shot area on a wafer, aligns them, and prints them.
It is considered essential to have a die-by-die exposure mode of the (through-the-reticle) alignment method.

Various methods have heretofore been considered for the TTR alignment method, but the most promising is that the reticle mark and the wafer mark can be simultaneously formed using exposure light and illumination light having a different wavelength. Another wavelength TTR to detect
This is an alignment method. In this method, the resist layer on the wafer can avoid the phenomenon of strongly absorbing exposure, so that stable mark detection is possible even for dye-containing resists and multi-layer resist wafers, and illumination at the time of alignment is possible. There is an advantage that exposure of the resist in the mark area can be prevented. As a typical technique (projection exposure apparatus) well known in another wavelength TTR alignment method, USP. 4,2 5 1,1 60 (Japanese Unexamined Patent Publication No.
No. 54369), USP. 4,2 6 9,5 0 5
-70025), USP. 4,4 9 2,4 5 9
No. 6-110234) or USP. 4,4 7 3,2 9
3 (JP-A-55-135831) and the like are known.

However, in these typical projection exposure apparatuses, a correction optical system for correcting chromatic aberration for alignment illumination light of another wavelength is interposed between the reticle and the projection lens. Such a correction optical system is used to maintain the reticle mark and the wafer mark in an imaging relationship with each other under illumination light of different wavelengths, but the alignment lacks stability or the accuracy is reproduced. The essential problem of not being able to obtain the character could not be solved.

[0006] Recently, a method for enabling another wavelength TTR alignment without using such an unstable and non-reproducible correction optical system has been disclosed in Japanese Patent Application Laid-Open Nos. 63-153820 or 63-153820. 63-283129. In the alignment system disclosed herein, a beam illuminating a reticle mark and a wafer mark is simultaneously imaged on each of two surfaces by a bifocal element and an objective lens, and one surface is a reticle pattern surface ( Mark surface), and the other surface is matched to a wafer conjugate plane in a space separated from the reticle pattern plane according to the axial chromatic aberration of the projection lens.

By employing such a method, it is not necessary to provide an optical element other than the projection lens in the alignment optical path between the reticle and the wafer, and TTR alignment can be performed in the same way as using exposure light. become. However, in the latest projection lens, various aberrations are satisfactorily corrected only for the exposure light, and both the axial chromatic aberration and the chromatic aberration of magnification occur for the wavelength shifted from the exposure light. Even if it is possible to cope with axial chromatic aberration, it is not always possible to cope with lateral chromatic aberration, and it is necessary to remove an alignment error (offset) caused by lateral chromatic aberration by some method. Therefore, techniques for coping with lateral chromatic aberration (distortion) are disclosed in, for example, JP-A-63-177421 and JP-A-63-2679.
No. 81427.

In Japanese Patent Application Laid-Open No. 63-177421, a reference mark is provided on a wafer stage which emits two light beams, ie, exposure light and alignment illumination light of a different wavelength. A technique is disclosed in which a distortion map in the field of view of a projection lens is created in advance by moving the image on the image plane side and scanning the back projection image of the light emitting mark on the reticle side to determine the position of the reticle mark. On the other hand, Japanese Unexamined Patent Publication (Kokai) No. 63-281427 discloses that an exposure light is converted into a beam and guided to an illumination beam transmitting system (scanner mirror, bifocal element, objective lens, etc.) of a different wavelength TTR alignment system. Difference in the amount of displacement detection for each wavelength beam when the illumination light beam and the illumination light beam of the exposure wavelength are simultaneously scanned and optical information from each of the reticle mark and the reference mark on the wafer stage is photoelectrically detected. Discloses a technique for knowing the distortion at the alignment point.

[0009]

In the prior art for coping with the chromatic aberration of magnification, the illumination light for alignment and the illumination light for exposure pass through the same position in the field of view of the projection lens. That is, JP-A-63-177421
In Japanese Patent Application Laid-Open No. 63-281427, it is necessary to guide two types of illumination light to the back side of the same light emitting reference mark via a fiber or the like. It is necessary to guide.

Such a structure complicates the configuration of the alignment optical system, and the performance of the optical element (particularly, achromatism).
In addition to manufacturing difficulties such as the need for strict requirements, the optical system conditions (jigma value, numerical aperture, telecentricity, etc.) under the illumination light for exposure, and the illumination for alignment There was a problem that it was difficult to match the optical system conditions under light, and it was difficult to know a practically accurate distortion error. Further, an actual exposure apparatus (stepper) can cope with exposure of a wide field (field of view) of about 15 × 15 mm to 20 × 20 mm even with a １ ／ reduction type. However, the exposure area of the reticle pattern used by the stepper user has various sizes, and accordingly, the position of the alignment mark in the projection visual field naturally changes variously. Since the amount of distortion of the projection lens with respect to the ideal grating under the illumination light for exposure also changes according to the change in the alignment position, the difference from the distortion characteristics under the illumination light for alignment is calculated only at the alignment position. There was also a problem that it was not enough to know only when superimposing the entire field.

Further, this is disclosed in the above-mentioned JP-A-63-2814.
Although this is a problem inherent in the device shown in Japanese Patent Publication No. 27,
When the exposure illumination light and the alignment illumination light are separated in a wavelength range by a dichroic mirror inclined at 45 ° above the reticle, the transmittance (or reflectance) of the dichroic mirror to the alignment illumination light is extremely high. If the height is increased, the illumination light for exposure to be detected by the different wavelength TTR alignment system hardly transmits (or reflects), and it becomes difficult to detect the mark.
Therefore, if the wavelength characteristic of the dichroic mirror is selected so that it has some transmittance (or reflectivity) for the exposure illumination light, the amount of light directed from the exposure illumination system to the reticle at the time of exposure is reduced by that amount. There was also a problem of doing.

According to the present invention, a mask pattern is accurately formed on a chip pattern formed on a sensitive substrate by using another wavelength alignment system for detecting and correcting an alignment offset caused by a projection optical system with high accuracy. It is an object of the present invention to provide a projection exposure apparatus and a method capable of superimposing and transferring images.

[0013]

In order to achieve this object, the present invention provides an illumination optical system for irradiating a mask (R) with exposure illumination light, and a projection optical system in which aberration is corrected for the exposure illumination light. (PL), and is applied to a projection exposure apparatus that transfers a pattern of a mask (R) onto a sensitive substrate (W) via a projection optical system (PL). In the present invention, the illumination light for detection (Lm1, Lm1) having a different wavelength from the illumination light for exposure is used.
2) is irradiated on the first reference mark (Fu), optical information (BT) generated from the first reference mark is detected by the first mark detection system (different wavelength alignment system AO2x), and the first information is detected.
The second reference mark (FMr) is irradiated with illumination light for exposure at a position different from the detection position of the first reference mark by the mark detection system, and the mark (RMr) on the mask (R) is projected via the projection optical system (PL). Is detected by a second mark detection system (exposure light alignment system AO1), and detection of the first reference mark by the first mark detection system and detection of the second reference mark by the second mark detection system are performed. A reference plate (FR) in which first and second reference marks are formed in a predetermined positional relationship is arranged on the photosensitive substrate side with respect to the projection optical system so that the projection can be performed simultaneously. After the adjustment of the optical characteristics of the projection optical system by the adjustment device (50A) or the adjustment of the position of the first mark detection system by the drive unit (10), the first and second positions are adjusted.
It is preferable to detect the reference mark. Furthermore, the first
The reference plate (FP) is used while the first and second reference marks (Fu, FMr) are detected by the first and second mark detection systems, respectively.
Is preferably kept substantially stationary. Here, another wavelength TTR alignment system is used as the first mark detection system of the present invention, and the principle of the present invention will be described with reference to FIG. FIG. 1 schematically shows a configuration of a stepper, and a circuit pattern area PA of a reticle R is formed on a wafer (not shown in FIG. 1) by a telecentric projection lens PL on both sides.
It is projected on the upper one shot area. Around the pattern area PA of the reticle R, marks Au and Al detected by TTR alignment systems AO2x and AO2y using illumination light of another wavelength (wavelength λ2) are provided.
The mark Au detected by the alignment system AO2x is used for alignment in the X direction, and the mark Al detected by the alignment system AO2y is used for alignment in the Y direction. And a side portion extending to the side. In addition, another wavelength TT
The R alignment systems AO2x and AO2y can move the alignment position in the X and Y directions. Further, outside the pattern area PA of the reticle R, a mark RMr detected by the TTR alignment system AO1 using the illumination light of the exposure wavelength λ1 is formed in the field of view of the projection lens PL. This mark RMr indicates that the pattern area PA
Reticles of different sizes are always fixedly arranged at a fixed location. The exposure light for uniformly illuminating the pattern area PA at the time of exposure is determined so as not to be shielded by the tip of the TTR alignment system AO1. Therefore, even when the reticle is replaced with a reticle having a different size of the pattern area PA, the TTR alignment system AO1 does not need to be moved and can be fixed on the apparatus.

On the other hand, below the projection lens PL, a reference mark plate FP fixed on the wafer stage is located. The surface of the reference plate (fiducial plate) FP is conjugate with the reticle R with respect to the projection lens PL under the exposure wavelength. On the surface of the reference plate FP, a reference mark plate FMr detected simultaneously with the mark RMr on the reticle R and reference marks Fu and Fl detected simultaneously with the marks Au and Al on the reticle R are etched with a chrome layer. And the like. The mark Fu corresponds to the mark Au and is provided with a large area so as to cover the entire movable position of the mark Au on the reticle R.
The same applies to the mark Fl.

When the reference plate FP is positioned such that the mark RMr and the reference mark FMr are simultaneously detected by the TTR alignment system AO1 as shown in FIG. 1, for example, the mark Au and the reference mark Fu are aligned with the TTR alignment system AO2x. At the same time. Since the positional relationship between the reference marks FMr and Fu on the reference plate FP and the positional relationship between the marks RMr and Au on the reticle R are known in advance, the alignment error detected by another wavelength TTR alignment system AO2x is used. The difference between the amount and the alignment error amount detected by the exposure wavelength TTR alignment system AO1 is the X-direction amount of offset caused by chromatic aberration at the position of the mark Au.

Therefore, if this offset amount is stored, it can be easily corrected later when the reticle R and the wafer W are aligned using the TTR alignment system AO2x. In this way, by making the separate wavelength alignment system and the exposure wavelength alignment system independent from each other and performing the mark detection operation at the same time, it is possible to more strictly correct the influence of distortion.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The construction of a stepper according to a preferred embodiment of the present invention will be described below with reference to FIGS. In this embodiment, the above-mentioned Japanese Patent Application Laid-Open No. Sho 63-28
As disclosed in Japanese Unexamined Patent Publication No. 3129, a two-beam interference type different wavelength TTR alignment system using a diffraction grating is used. However, other types (image detection system, spot scan system) are similarly used. Applicable.

In FIG. 2, a dichroic mirror DM is inclined at 45 ° above the reticle R to bend the optical axis AX of the projection lens PL horizontally. The dichroic mirror DM reflects exposure light, which travels from an exposure illumination system (not shown) along the optical axis AX, by approximately 90% or more, and directs the light to the pattern area PA of the reticle R. Reticle R is 2
The reticle stage RS is held on a reticle stage RS that moves minutely in the dimension (X, Y, θ directions), and the reticle stage RS is positioned by the driving unit 1. On the other hand, the wafer W and the reference plate FP are held below the projection lens PL. An XY stage ST is provided, is moved in the X and Y directions by the motor 2, and the coordinate position of the stage ST is sequentially measured by the laser interferometer 3. The positioning of the stage ST is performed by monitoring the measured value of the interferometer 3. This is performed by a stage driver circuit 4 that drives the motor 2. The driver circuit 4 controls the movement and positioning of the stage ST based on a command from a main control device 5. The main control device 5
Also manage the control of

The different wavelength TTR alignment system of this embodiment is constructed with two to four eyes (the number of objective lenses) above the dichroic mirror DM.
Only the eye, here the system corresponding to AO2x in FIG. 1, is shown.
The telecentric objective lens OBJu and the mirror M1 are held by hardware 11 and moved in the X and Y directions by the drive unit 10.

The objective lens OBJu is arranged so as not to interfere with the dichroic mirror DM and so that its optical axis is perpendicular to the reticle R. Illumination light for different wavelength alignment is emitted as a linearly polarized beam LB from a laser light source 12 such as helium-neon or argon ion, and enters a two-beam frequency shifter unit 13. In the unit 13, the beam LB is split into two,
Two AOMs (acousto-optic modulators) 130 that apply high-frequency modulation (frequency shift) to each of the two split beams.
A, 130B and small lenses 131A, 131B for converging the frequency-shifted beams are provided. A
The OM 130A is driven at a frequency f1 (for example, 80 MHz), and the first-order diffracted beam LB1 is
AOM 130B is extracted via frequency A2
(For example, 80.03 MHz), and the first-order diffraction beam LB2 is extracted through the small lens 131B. The principal rays of the two beams LB1 and LB2 are determined in parallel and symmetrically with the optical axis of the alignment system, and are split into two by the beam splitter 14. The split beams Lr1 and Lr2 that have passed through the beam splitter 14 are converted by a condenser lens (inverse Fourier transform lens) 15 into parallel light beams that intersect at the rear focal plane. A reference grating 16 is arranged on the rear focal plane, where two beams Lr1,
Interference fringes having a pitch corresponding to the crossing angle of Lr2 and the wavelength are formed. This interference fringe is a frequency difference Δ between the two beams Lr1 and Lr2.
It moves in one direction on the reference grating 16 at a speed corresponding to f (30 KHz). On the reference grating 16, a transmission type diffraction grating having a pitch equal to the pitch of the interference fringes is provided in parallel with the fringes. Accordingly, from the reference grating 16, the interference light DR1 of the zero-order light of the beam Lr1 traveling in the same direction as the beam Lr1 and the primary light generated from the beam Lr2, and the zero-order light of the beam Lr2 traveling in the same direction as the beam Lr2 And an interference light DR2 with the primary light generated from the beam Lr1 are generated, and the photoelectric element 17 detects the amounts of these interference lights DR1 and DR2. Where the interference light DR
1 and DR2 change in intensity in a sinusoidal manner at the frequency difference Δf (30 KHz), that is, at the beat frequency, and the output signal SR from the photoelectric element 17 becomes an AC signal of 30 KHz. This signal SR serves as a reference signal for phase comparison during alignment.

FIG. 3 shows the beams LB1, LB2 and the lens 1
FIG. 4 is an optical path diagram showing in detail the state of each beam from No. 5 to the photoelectric element 17. First, two beams LB1 and LB2 are respectively condensed on the surface EPa by the action of the small lenses 131A and 131B to become a beam waist. The surface EPa is a Fourier surface coinciding with the front focal plane of the lens 15, and the beams Lr 1 and Lr 2 form parallel light beams from the lens 15 and intersect on the reference grating 16. The positions of the beams Lr1 and Lr2 within the plane EPa are determined symmetrically with respect to the optical axis AXa of the alignment system, and the distance from the optical axis AXa is determined by the angle of incidence θ of the beams Lr1 and Lr2 with respect to the reference grating 16. decide.

According to a general analysis theory, the wavelength λ
The diffraction angle α of the ± 1st-order diffracted light with respect to the 0th-order light, which is determined by the coherent light of
It is uniquely determined by sinα = λ / Pg. Therefore, when the grating pitch Pg of the reference grating 16 is set so that the diffraction angle α of the first-order diffracted light becomes exactly 2θ with respect to the 0-order light (interference light DR1) of the beam Lr1, the first-order diffracted light becomes 0% of the beam Lr2. It is generated coaxially with the next light (interference light DR2).

At this time, the pitch of the interference fringes formed on the reference grating 16 is equal to the grating pitch Pg. The photoelectric element 17 has light receiving surfaces 170A and 170B for individually receiving the two interference lights DR1 and DR2.
The photoelectric signal from 170B is added in an analog manner by adder 171 and output as reference signal SR.

Returning to the description of FIG. 2, the two beams Lm1 and Lm2 reflected by the beam splitter 14 are:
Further, the light is reflected by the beam splitter 18 and enters the objective lens OBJu via the mirror M1. Objective lens OBJ
u converts the two beams Lm1 and Lm2 into parallel light beams that intersect at a plane IP in space. The surface IP is separated from the reticle R by the amount of axial chromatic aberration ΔL in the optical axis AX direction, and is a surface conjugate with the wafer W or the reference plate FP under the wavelength (λ2) of the alignment illumination beam LB. The two beams Lm1 and Lm2 intersecting at the plane IP are separated from each other in the mark area Au on the reticle R, and pass through the pupil E of the projection lens PL.
After the beam waist at P, the beam crosses again on the reference plate FP (or wafer W) as a parallel light beam.

FIG. 4 is an optical path diagram showing the state of the beams Lm1 and Lm2. The front focal plane of the objective lens OBJu coincides with the plane EPa shown in FIG. 3, and the two beams Lm1 and Lm2
Becomes the beam waist there. The two beams Lm1 and Lm2 emitted from the objective lens OBJu both become parallel light beams and intersect at the plane IP, that is, at the rear focal plane of the objective lens OBJu. Thereafter, the beam Lm1 irradiates the reticle grating mark Aua in the mark area Au of the reticle R, passes through the transparent part, and enters the projection lens PL, and becomes a parallel light beam again on the reference plate FP (or wafer W). The grid in the mark area Fu is irradiated obliquely.

Similarly, the beam Lm2 irradiates the reticle grating mark Aub on the reticle R, passes through the transparent portion, and passes through the mark area F on the reference plate FP (or wafer W).
The grating in u is irradiated obliquely from a direction symmetric to the beam Lm1. As shown in FIG. 4, the two beams Lm1, Lm
m2 is the pupil space (object lens OBJu and small lens 131A,
131B, and inside the projection lens PL), the light beam is a convergent light beam and the principal ray is parallel to the optical axis AXa,
Image space (between objective lens OBJu and projection lens PL and between projection lens PL and reference plate FP or wafer W)
In this case, the light beams are parallel light beams and the principal rays intersect at the focal plane (image plane).

From the mark area Fu on the reference plate FP, interference light BT is generated in the vertical direction. The interference light BT is vertically generated from the mark Fu by the irradiation of the beam Lm1.
The second-order diffracted light interferes with the first-order diffracted light generated perpendicularly from the mark Fu by the irradiation of the beam Lm2, and is a parallel light beam in the image space. This interference light BT is projected on the projection lens P
After converging on the beam waist at the center of the pupil EP of L, the transparent portion (mark Aua) in the center of the mark area of the reticle R
And Aub) as a parallel light beam, and travels backward along the optical axis AXa of the objective lens OBJu. The interference light BT converges again on the beam waist at the center of the front focal plane of the objective lens OBJu, that is, the center of the plane EPa conjugate with the pupil EP of the projection lens PL, and the mirror M1 and the beam splitter 18 in FIG. Return to Here, the structure of the mark area Au on the reticle R will be described with reference to FIGS. As shown in FIG. 7, a light-shielding band ESB having a fixed width is formed on the outer periphery of the circuit pattern area PA of the reticle R.
A transparent window is provided in part of the window, and two lattice marks Au
a and Aub are arranged with the light-shielding portion ESB 'interposed therebetween. The pitches of the marks Aua and Aub are equal, and the width in the direction orthogonal to the pitch direction is set to about half of the transparent window. The width of the light-shielding portion ESB 'is also set to about half of the transparent window, and the lattice mark WMu on the wafer W is located at the portion of the mark region WMU' under the wavelength for alignment. The two beams Lm1 and Lm2 respectively illuminate each of the lattice marks Aua and Aub in a rectangular shape, but the beams Lm1 that have passed through the transparent portion adjacent to the mark Aua on the pattern area PA side.
Irradiates the wafer mark WMu as shown in FIG. 8, and the beam Lm2 that has passed through the transparent portion of the mark Aub adjacent to the pattern area PA side irradiates the wafer WMu. This mark W
The interference light BT generated vertically from Mu returns to the objective lens OBJu on the reticle R through the portion WMu 'in FIG.

In FIG. 4, when the beams Lm1 and Lm2 irradiate the reticle grating marks Aua and Aub, respectively, the higher-order diffracted light is reflected together with the zero-order light D01 and D02. In this embodiment, the grating mark A of the beam Lm1 (Lm2)
ua (Aub), the angle of incidence of the first-order diffracted light is 0
The pitch of the mark Aua (Aub) is determined so that the diffraction angle with respect to the next light D01 (D02) is exactly 2θ '. Thereby, the first-order diffracted light D11 (frequency shift f1) that travels backward from the grating mark Aua along the optical path of the beam Lm1 as it is.
Is generated, and a first-order diffracted light D12 (frequency shift f2) that travels backward from the grating mark Aub along the optical path of the beam Lm2 as it is.
Occurs. Therefore, these first-order diffracted lights D11 and D12 also return to the light receiving system via the objective lens OBJu together with the interference light BT.

Referring again to FIG. 2, the interference light BT and the first-order diffracted lights D11 and D12 pass through the beam splitter 18 and enter the condenser lens system 19. The lens system 19 is also an inverse Fourier transform lens, and includes the interference light BT and the first-order diffracted lights D11 and D11.
Both 12 are converted into parallel light beams, and they are crossed at the focal plane (image conjugate plane). Each beam having passed through the lens system 19 is split into two beams by a beam splitter 20, and the transmitted beam reaches a reticle reference grating plate 21, and the reflected beam reaches a field stop 22. Both the reference grating plate 21 and the field stop 22 are provided with a plane IP (reference plate FP or wafer W).
Conjugate plane). Therefore, in the reference grating plate 21, the first-order diffracted lights D11 and D12 intersect, and an interference fringe is formed in the intersection area. This interference fringe naturally has a beat frequency Δ
It flows one-dimensionally at f (30 KHz). So Figure 6
As shown in (A), reference grid plate 2 covered with a chrome layer
1, a transmission type diffraction grating 21A is provided on this grating 21A.
The optical element 23 receives the interference light BTr of the diffracted light generated coaxially from the light. This will be described in detail with reference to FIG. Since the reference grating plate 21 is image conjugate, it is conjugate with the plane IP and the wafer mark WMU (or the surface of the reference plate FP),
The interference light BT also enters together with the first-order diffracted lights D11 and D12. However, since the wafer mark WMu and the reticle marks Aua and Aub are laterally displaced in the XY plane, as shown in FIG. Grid 21 where D12 intersects
The interference light BT returns to the portion (light blocking portion) 21B beside A.

Therefore, the grating 21A on the reference grating plate 21
The interference light BT from the wafer mark WMU can be shielded only by matching the position and size of the mark with the size of the mark Aua or Aub. Each zero-order light BT of the first-order diffracted lights D11 and D12 illuminating the grating 21A of the reference grating plate 21
o advances so as to deviate from the photoelectric element 23, and the grating 21A
Only the interference light BTr of the first-order diffracted light vertically generated from the light is received by the photoelectric element 23. This method is the same as the method of obtaining the interference light BT from the wafer mark WMu shown in FIG. 4, and in this case, the wafer mark WMu or the grating 2
The 1A grating pitch is exactly twice the pitch of the interference fringes formed thereon.

The intensity of the interference light BTr thus received by the photoelectric element 23 changes sinusoidally at a beat frequency Δf (30 KHz), and the output signal Sm of the photoelectric element 23 is a reticle mark Aua based on the reference grating 16. , Aub
Is an AC signal whose phase difference linearly changes with respect to the reference signal SR in accordance with the amount of displacement in the pitch direction. On the other hand, the interference light BT and the first-order diffracted lights D11 and D12 reflected by the beam splitter 20 shown in FIG. 2 reach the field stop 22, which stops only the interference light BT as shown in FIG. An opening 22A through which the first order diffracted light D
11 and D12 are shielded by the light shielding portion 22B.

The interference light BT that has passed through the field stop 22 reaches the photoelectric element 26 via the mirror 24 and the condenser lens 25. The photoelectric element 26 generates an output signal Sw corresponding to a change in the intensity of the interference light BT. This signal Sw also has a beat frequency Δf
The signal becomes an AC signal whose level changes sinusoidally in the
Upper mark WMu or reference mark Fu on reference plate FP
Of the reference signal SR in proportion to the amount of displacement of the reference
Changes with respect to.

The above reference signal SR and output signals Sm, Sw
Are input to the phase difference measuring unit 27, and the measuring unit 27 outputs the phase difference φ of the output signal Sm with respect to the reference signal SR.
m and the phase difference φw between the output signal Sw and the reference signal SR.
And the difference Δφ = φm−φw is calculated.
In the case of this embodiment, since the pitch of the diffraction grating on the reference plate FP (or the wafer W) is twice as large as the pitch of the interference fringes formed thereon, one cycle (± 180 °) of the phase difference Δφ is This corresponds to ピ ッ チ of the diffraction grating pitch (± １ ／ pitch). Based on this phase difference Δφ, measurement unit 27 calculates position correction amounts (position shift amounts) ΔX and ΔY of wafer stage ST or reticle stage RS, and sends the values to main controller 5.

As described above, the configuration from the drive section 10 to the measurement unit 27 including the objective lens OBJu corresponds to the first mark detection means of the present invention. In FIG. 2, the mark RMr on the reticle R is indicated by a mirror M2 and an objective lens OB.
Exposure light TTR alignment system (shown in FIG. 1) including a beam splitter 30, a lens system 31, an illumination field stop 32, a condenser lens 33, a fiber 34, an imaging lens 35, a beam splitter 36, and CCD image sensors 37A and 37B. AO1). The fiber 34 emits illumination light having an exposure wavelength and uniformly irradiates the diaphragm 32 via a condenser lens 33. The illumination light passing through the aperture of the stop 32 enters the objective lens OBr via the lens system 31 and the beam splitter 30, is further bent vertically by the mirror M2, and irradiates a local area including the mark RMr of the reticle R. Here, the stop 32 is conjugate with the reticle R, and an aperture image of the stop 32 is formed on the reticle R. The beam splitter 30 is disposed near the front focal plane of the telecentric objective lens OBr, that is, in the vicinity of a plane conjugate with the pupil EP of the projection lens PL, and forms a part of light returning from the objective lens OBr into an imaging lens 35. Reflects toward. CCD image sensor 37A, 3
The light receiving surface of 7B is conjugate with the reticle R by the objective lens OBr and the imaging lens 35, and at the same time, the projection lens PL
And the wafer W or the reference plate FP.
The exit end of the fiber 34 is the pupil EP of the projection lens PL.
And Koehler illumination is performed. In the case of this embodiment, an afocal system is provided between the objective lens OBr and the beam splitter 30, and the objective lens OBr
It is possible to cope with the position polarization of the mark RMr by integrally moving the mirror M2 and the mirror M2 in the horizontal direction in FIG. 2, but here, the mirror M2 is taken into consideration in consideration of the stability of the system. The reticle R is fixedly arranged so as to be located outside the maximum dimension of the pattern area PA assumed on the reticle R.

The CCD elements 37A and 37B are rotated by 90 ° with respect to each other so that the horizontal scanning lines are in the X direction and the Y direction, respectively. The position measurement in the direction is performed separately. This is because ordinary CCD elements have different pixel resolutions in the horizontal and vertical directions, so that a single CCD element avoids a difference in resolution when detecting a shift in both the horizontal and vertical mark images. is there. The image processing unit 38
An image signal (video signal) is input from each of the CCD elements 37A and 37B, and a reference mark FMr on the reference plate FP is input.
And a position shift amount between the mark RMr on the reticle R and the mark RMr on the reticle R.

In addition to the above configuration, an off-axis type global mark detection system 40 for detecting a global alignment mark on the wafer W, a latent image in the resist layer, or each reference mark on the reference plate FP, A processing unit 42 is provided. Also, an adjustment unit 5 for adjusting or correcting various image forming characteristics of the projection lens PL.
0A and 50B are provided, and are integrally controlled by the main control device 5. The adjustment unit 50A includes, for example, a projection lens P
L has a function of changing the magnification, focal position, distortion and the like of the projection lens PL by controlling the pressure of a predetermined air chamber in the projection lens PL.
Has a function of slightly moving or tilting some of the lens elements (for example, a reticle-side field lens) in the optical axis direction.

The different wavelength TTR alignment system and the exposure light TTR alignment system are preferably arranged as schematically shown in FIG. Like the reticle R shown in FIG. 9, each side of the light-shielding band ESB around the pattern area PA is provided with a reticle lattice mark area Au, as shown in FIG.
Al, Ar, and Ad are formed. Mark area Au, Ad
Are used for alignment in the X direction in FIG. 9, and the mark areas Ar and Al are used for alignment in the Y direction.

Further, a similar reticle mark RM is provided at a position symmetrical to the reticle mark RMr outside the light-shielding band ESB.
1 is provided. Accordingly, the objective lenses OBr and OBl of the two exposure light TTR alignment systems are arranged so as to detect the marks RMr and RM1 under the dichroic mirror DM, and the objective lenses OBJu, OBJd, OBJr and OBJr of the four different wavelength TTR alignment systems are arranged. OBJl are respectively connected to the mark areas Au, via the dichroic mirror DM.
It is arranged to detect Ad, Ar, Al.

When the reticle R and the wafer W are actually die-by-die aligned using a different wavelength TTR alignment system, it is not necessary to use four eyes at the same time. You may. This corresponds to the mark (WMu, WM) for one shot area on the corresponding wafer.
d, WMr, WMl), even if there is a defective portion, if at least one X-direction eye and one Y-direction eye properly output a photoelectric signal, alignment of the shot is performed. Is executed, alignment errors (interruption of the sequence) can be avoided as much as possible.

Next, the mark arrangement of the reticle R shown in FIG. 9 and the mark arrangement on the reference plate FP suitable for this embodiment will be described with reference to FIGS. FIG. 10 shows a pattern arrangement of a preferred example of the reticle R. In this case, a pattern area P having the largest dimension that can be projected by the projection lens PL is shown.
Suppose A. Center R of pattern area PA of reticle R
Two cross-shaped reticle marks RMr and RMl are provided on a line segment passing through cc and parallel to the X axis. These marks RMr and RMl are set substantially at the centers of the fields of view of the objective lenses OBr and OBI, respectively. In the reticle R shown in FIG. 10, four mark areas Au, Ad, Ar,
Al are both located farthest from the reticle center Rcc, and are rectangular regions Su, S indicated by broken lines in FIG.
d, Sr, and Sl are objective lenses OBJu, OBJd, OBJr, and OBJ of different wavelength TTR alignment systems, respectively.
1 shows an example of a movable range of an optical axis (detection center) of l. FIG.
In the case of, the mark areas Au, Ad, Ar, and Al are all provided at the outermost positions of the movable range of the objective lens.

FIG. 10 is an example. In some cases, the movable ranges Sr, S of the objective lenses OBJr, OBJl are located between the movable ranges Su, Sd of the objective lenses OBJu, OBJd.
l may enter from the left and right.
FIG. 11 shows the arrangement of each mark on the reference plate FP.
Reference marks FMr, FMl in the form of double crosses on the left and right of the direction
Is provided. The distance between the centers of the two marks FMr and FMl is determined to be equal to the value obtained by multiplying the distance between the centers of the marks RMr and RM1 on the reticle by the projection magnification (1 / M). Therefore, the center Fcc of the reference plate FP is set to the reticle center Rc.
When the reticle mark RMr is positioned between the double lines of the mark FMr, they are simultaneously observed by the objective lens OBr of the exposure light TTR alignment system, and when the reticle mark RMr is located between the double lines of the mark FMr. With the reticle marks RMl located in between, they are simultaneously observed by the objective lens OBl.

Further, on the reference plate FP, the center Fcc
And the reticle center Rcc, the reticle R
Reference mark areas Fu, Fd, Fr, and Fl are formed by engraving the diffraction grating so as to have positions and sizes corresponding to the upper movable ranges Su, Sd, Sr, and Sl, respectively. For example, in the reference mark area Fu, the mark WMu on the wafer W
In the same manner as described above, a group of gratings engraved at a constant pitch in the X direction is formed, and is used to detect a relative positional deviation in the X direction from the mark area Au on the reticle R. Other fiducial mark area F
The same applies to d, Fr, and Fl. Therefore, even if the size of the pattern area PA (light-shielding band ESB) changes due to the exchange of the reticle R, each mark area A on the reticle can be changed.
u, Ad, Ar, and Al are movable ranges Su, Sd, Sr, S
l, the mark area Au and the reference mark Fu are displaced in the X direction, and the mark area Ad and the reference mark Fd
Of the mark area Ar and the reference mark Fr
Direction deviation, and Y of the mark area Al and the reference mark Fl
The displacement of the direction is changed by four eyes (OBJu, OBJd, OBJr,
OBJl) can be detected simultaneously.

In FIG. 11, the left and right reference marks Fr
The plurality of gratings forming F1 and Fl correspond one-to-one with respect to the Y direction, and the plurality of grids forming the upper and lower reference marks Fu and Fd correspond one-to-one with respect to the X direction. The center distance in the X direction between the mark areas Au and Ad shown in FIG. 10 is exactly separated by an integral multiple of the grid pitch of the reference marks Fu and Fd, and the center distance in the Y direction between the mark areas Ar and Al is the reference mark. It is set so as to be accurately separated by an integral multiple of the grating pitch of Fr and Fl.

Next, the operation of this embodiment, that is, the baseline measurement between the exposure light TTR alignment system and the different wavelength TTR alignment system in consideration of distortion will be described. First, an arbitrary reticle (for example, FIG. 10) is set on the reticle stage RS, and the reticle marks RMr and RMl and the marks FMr and RMr of the reference plate FP are set using the CCD elements 37A and 37B of the exposure light TTR alignment system.
The reticle alignment is performed by imaging the FM1.

Next, objective lenses OBJu, OBJd, OBJr, OBJl of four different wavelength TTR alignment systems
(Hardware 11) in each mark area Au, Ad, Ar, Al
, The deviation of the irradiation position of the two beams Lm1 and Lm2 in each direction and the two beams Lm1 in each direction using the reference marks Fu, Fd, Fr and Fl on the reference plate FP. , Lm2 are checked. When the check is completed, four different wavelength TTRs at that position
Each of the alignment systems obtains a relative position shift amount between the reticle R and the reference plate FP. That is, the measuring unit measures the amount of displacement in the X direction between the reference mark Fu (and Fd) and the reticle mark Au (and Ad) and the amount of displacement in the Y direction between the reference mark Fr (and Fl) and the reticle mark Ar (and Al). 27. Main controller 5 controls drive unit 1 or stage driver 4 based on the detected deviation amount to servo drive reticle stage RS or wafer stage ST. Since the different wavelength TTR alignment system of the present embodiment is a heterodyne system capable of sequentially measuring the relative position shift amount even when the reticle mark and the reference mark are stationary, the measurement unit 27 sequentially sets the position shift amount. (Shifts in the X and Y directions and the amount of displacement in the rotation direction) are continuously output. Therefore, the reticle R and the reference plate FP are kept aligned so that all the phase differences Δφ measured by the measurement units 27 of the four different wavelength TTR alignment systems become zero (or a constant value). During this different wavelength TTR alignment, the image processing unit 38 of the exposure light TTR alignment system adjusts the positional deviation amounts (ΔXr, ΔYr) between the reticle mark RMr and the reference mark FMr in the X and Y directions, and the reticle mark RM.
1 and the reference mark FMl in the X and Y directions (Δ
Xl, ΔYl) are successively output (at fixed time intervals). However, when the distortion is large, the deviation amount due to the distortion is added in advance as the offset amount by the design value. The deviation amounts (ΔXr, ΔYr) and (ΔXl, ΔYl) obtained by the exposure light TTR alignment system in this way are:
Alignment position between another wavelength and exposure light wavelength (mark A
u, Ad, Ar, Al). Here, more specifically, the reticle mark Au (reference mark F
u) and Ad (reference mark Fd) are detected,
When both of the phase differences Δφ are zero, the positional deviation amounts (ΔXr, ΔY
r) and (ΔXl, ΔYl) are stored a plurality of times, and the stored results are averaged to obtain a total alignment error (X, Y, θ directions) due to a distortion difference between the reticle R and the reference plate FP. ) Is required. In the case of this embodiment, since the reticle R and the reference plate FP are simultaneously aligned with the four eyes of the different wavelength TTR alignment system, as a result of the alignment, the reticle R and the reference plate FP are distorted by different wavelengths of the projection lens PL. Depending on the characteristics, there are slight errors in the X, Y, and θ directions. This error is due to the reticle marks Au, Ad, Ar, A
As long as the position of 1 does not change, it can be considered as a constant offset. Therefore, the mark RM is used in the exposure light TTR alignment system.
When the shift amount between r, RMl and the reference marks FMr, FMl is detected, the shift amount includes X, Y, θ including the amount of distortion at the position of the marks RMr, RMl of the projection lens under the exposure light. The offset amount in the direction is measured.

Therefore, when the shots on the wafer W are actually aligned by another wavelength TTR alignment system thereafter, the reticle stage RS or the wafer stage ST is set such that the position shifted by the offset amount becomes the true alignment achievement position. May be controlled. X, Y,
The offset amount due to the distortion difference in the θ (rotation) direction is calculated in main controller 5 based on the deviation amount information from unit 38, and is stored until reticle R is realigned or replaced.

Note that the distortion difference between the different wavelength and the exposure wavelength in the projection lens PL can change when the adjustment unit 50B is driven. Therefore, immediately after the unit 50B is largely driven, the distortion difference is again determined using the reference plate FP. It is desirable to measure the offset amount. As described above, according to the present embodiment, the mark regions Au, Ad, Ar, and Al for different wavelength TTR alignment are provided on the four sides on the reticle R, and the reference plate F
Since the corresponding reference marks Fu, Fd, Fr, and Fl on P are simultaneously detected, the total alignment error in the TTR due to the influence of distortion due to another wavelength can be accurately obtained. Further, in this embodiment, the different wavelength TTR alignment system and the exposure light TTR alignment system are simultaneously operated via the projection lens PL, and the measured values of the interferometer 3 of the wafer stage ST are not used at all. Accordingly, an error due to air fluctuation in the atmosphere, which poses a problem, does not intervene, and extremely accurate offset amount measurement can be performed.

Also, in this embodiment, since a different wavelength TTR alignment system of a very high resolution heterodyne system is used, for example, the pitch of the diffraction grating on the reference plate FP is set to 4 μm.
The phase difference detection range (± 18 °) is ± 1 μm
If the practical phase measurement resolution considering noise etc. is ± 2 °, the displacement detection resolution is approximately ± 0.01μ.
m.

Therefore, another wavelength TTR of the heterodyne system
When the alignment servo between the reticle R and the reference plate FP is performed using the alignment system, extremely stable positioning is achieved. By the way, the projection lens PL of this embodiment is telecentric on both sides, but may be of course one-sided telecentric. In the case of a double-sided telecentric projection lens, the optical axis of the objective lens of the exposure light TTR alignment system is perpendicular to the reticle surface and the projection lens P
It also coincides with the principal ray passing through the pupil center point of L. For this reason, if the mark pattern on the reticle is made of a reflective chrome layer, the specularly reflected light from this pattern may be strongly detected by the CCD element, and the reticle mark RM
r, RMl and the reference marks FMr, FMl may both be bright. Also, a reference plate F made of quartz glass or the like is used.
When the entire surface of P is covered with a chrome layer and the reference marks FMr and FMl are cut out by etching or the like, the reference mark F
Although Mr and FMl look black, their surroundings may all be bright, and the contrast of the reticle marks RMr and RMl themselves may be extremely reduced. In such a case, an aperture stop (spatial filter) having an annular aperture is provided in the illumination light path of the exposure light TTR alignment system, for example, on the pupil conjugate plane between the beam splitter 30 and the lens system 31 in FIG. The reticle R may be provided for dark field illumination. At this time, a dark field image in which only the edges of the reticle marks RMr and RMl and the reference marks FMr and FMl shine brightly is formed on the CCD element. The spatial filter provided on the pupil conjugate plane is composed of liquid crystal or electrochromic (EC) patterned concentrically in an annular shape, switching between dark-field illumination and bright-field illumination, and changing the numerical aperture of illumination light. You can also do.

Next, a description will be given of a method of improving overlay accuracy using the apparatus shown in FIG. FIG. 12 shows the distortion characteristic (dashed line) under the exposure wavelength of the projection lens PL and the distortion characteristic (dashed-dotted line) under another wavelength (alignment light wavelength) based on the ideal lattice (solid line). It is exaggerated.

Such a distortion map can be created by performing trial printing using, for example, a test reticle. However, a distortion map at another wavelength cannot be created by trial printing, but can be created by using a method similar to that of the above-described embodiment. First, a test reticle having a vernier mark for overlay measurement at each ideal lattice point in the pattern area is prepared, and the test reticle is aligned by an exposure light TTR alignment system. Then, the reticle blind inside the exposure illumination system is fully opened, and the test reticle is exposed on a dummy wafer (a bare silicon wafer coated with a photosensitive resist layer, a photochromic layer, a magneto-optical medium, etc.).

Next, the reticle blind is closed so that only the vernier mark provided at the center of the test reticle is illuminated, and the wafer stage ST is stepped at the pitch of the ideal lattice point. Exposure is performed by overlapping with the vernier mark latent image. In this case, it is considered that the stepping of the wafer stage ST is coincident with the division pitch of the ideal lattice, and therefore, the latent image of each vernier mark exposed first and the latent image of the vernier mark superimposed for the second time are compared. By measuring the overlay accuracy, a distortion characteristic at the exposure wavelength with reference to the ideal grating is obtained. In this measurement, the latent image of the vernier mark on the dummy wafer may be detected by the off-axis type global mark detection system 40 shown in FIG.
It may be detected by an alignment system. In the case of a dummy wafer using a normal photoresist layer, once development is performed to create a vernier mark resist image,
The measurement may be performed by another wavelength TTR alignment system or the like.

As described above, the overlay error amount at each of the ideal lattice points is obtained, and each error amount is statistically processed by the least-squares method to obtain the offsets in the X, Y, and θ directions. Further, from the offset, the image R at the exposure wavelength of the reticle marks RMr and RMl indicated by the broken lines in FIG.
The amount of deviation (Δ) of Mr ′ and RMl ′ from the ideal position (solid line)
OFx1, ΔOFy1) and (ΔOFx2, ΔOFy2) can be estimated as system offsets.

Therefore, when aligning reticle marks RMr and RMl with the exposure light TTR alignment system,
By taking the system offset into account, exposure can always be performed on a distortion map with the least error from the ideal grating. Furthermore, the difference between the distortion at the exposure wavelength and the distortion at another wavelength is
Since the difference can be obtained by the embodiment, if the difference is further added (corrected), the pattern of the reticle can be obtained in the state closest to the ideal grating even in the case of die-by-die alignment using another wavelength TTR alignment system. The shot area on the wafer is overlaid and exposed. This means
This means that matching between a plurality of steppers constituting one semiconductor manufacturing line can be made based on an ideal lattice, and the matching accuracy within that line is improved.

In the above embodiment, it is assumed that the stepper shown in FIG. 2 is used.
The detection center of the TR alignment system is at a fixed position within the field of view of the projection lens. Therefore, as described with reference to FIG. 12, the system offsets (ΔOFx1, ΔOFy1) and (ΔOFx) during reticle alignment are set so that the distortion map at the exposure wavelength is closest to the ideal grating in advance.
2, ΔOFy2), and after correcting it, a distortion error of another wavelength at the alignment position may be obtained by using another wavelength TTR alignment system.

[0056]

As described above, according to the present invention, a different wavelength alignment system that uses an illumination light having a wavelength different from that of exposure light and detects an alignment mark on a sensitive substrate via a projection optical system. The baseline amount, that is, the offset amount generated at the time of alignment between the mask and the sensitive substrate using another wavelength alignment system can be accurately detected and corrected easily. In particular, when the separate wavelength alignment system and the exposure wavelength alignment system are separated so that the fiducial marks arranged on the image plane of the projection optical system are simultaneously detected, the distortion difference when using another wavelength can be accurately determined. It becomes possible to measure. That is, the alignment light of the exposure wavelength and the alignment light of another wavelength via the projection optical system are detected by the respective alignment systems at exactly the same time. For this reason, measurement errors due to subtle air fluctuations in the optical path, fluctuations in the temperature distribution, and the like are commonly included and can be canceled. Also, while the reference mark is being detected by another wavelength alignment system and the exposure wavelength alignment system, the reference plate is stationary, and it is not a method of relying on a laser interferometer while running a wafer stage or the like. There is also an advantage that the measurement accuracy (reproducibility) of the laser interferometer itself is not affected.

Also, in the present invention, a plurality of different wavelength alignment systems can be operated simultaneously using the reference mark, so that beam positioning, telecentric check, and focusing accompanying movement of the objective lens of the different wavelength alignment system can be performed. The check can be performed at the same time, and there is an advantage that the throughput of setting (adjustment) of another wavelength alignment system is improved.

[Brief description of the drawings]

FIG. 1 is a perspective view showing a principle configuration of a stepper according to an embodiment of the present invention.

FIG. 2 is a diagram showing a detailed configuration of a stepper according to an embodiment of the present invention.

FIG. 3 is a diagram illustrating an optical system of another wavelength TTR alignment system.

FIG. 4 is a diagram illustrating an optical system of another wavelength TTR alignment system.

FIG. 5 is a diagram illustrating an optical system of another wavelength TTR alignment system.

FIG. 6A is a plan view showing a reticle reference grating.
(B) is a top view which shows the field stop for wafers.

FIG. 7 is a plan view showing the arrangement of reticle grating marks.

FIG. 8 is a plan view showing the arrangement of wafer lattice marks.

FIG. 9 shows another wavelength TTR alignment system and exposure wavelength TTR.
FIG. 3 is a perspective view showing a practical arrangement with an alignment system.

FIG. 10 is a plan view showing a practical reticle pattern arrangement.

FIG. 11 is a plan view showing a pattern arrangement on a reference mark plate.

FIG. 12 is a diagram illustrating distortion characteristics due to exposure light and different wavelengths.

[Explanation of symbols]

R: reticle, RMr, RMl: reticle mark, PL: projection lens, OBJu, OBJd, OB
Jr, OBJl: Objective lens of another wavelength TTR alignment system, Au, Ad, Ar, Al: Different wavelength TTR
Reticle mark for alignment, OBr, OBI ...
・ Exposure light TTR alignment system objective lens, FP
・ Reference plate, ST: wafer stage, FMr, FMl
... Reference mark for exposure light TTR alignment, F
u, Fd, Fr, Fl... Reference marks for different wavelength TTR alignment

Claims (10)

(57) [Claims] 1. An illumination optical system for irradiating a mask with exposure illumination light, and a projection optical system corrected for aberrations with respect to the exposure illumination light, wherein the pattern of the mask is responsive to the substrate via the projection optical system. In a projection exposure apparatus for transferring the light onto the first reference mark, a first mark detection system that irradiates a first reference mark with detection illumination light having a different wavelength from the exposure illumination light, and detects light information generated from the first reference mark, Detection of the first reference mark by the first mark detection system
At a position different from the position, the exposure illumination light is
And irradiates the mask on the mask through the projection optical system.
A second mark detection system for detecting relative position information with respect to the mark, and detection of the first fiducial mark by the first mark detection system
The second reference mark by the second mark detection system.
In so that Do feasible detection and at the same time, the projection optical system
With respect to the sensitive substrate side, and the first and second
A projection exposure apparatus comprising: a reference plate on which a reference mark is formed in a predetermined positional relationship. 2. The method according to claim 1, wherein the first mark detection system is provided on the sensitive substrate.
Position information obtained by detecting the alignment mark
2. The projection exposure apparatus according to claim 1, wherein the projection exposure apparatus is used for alignment with the mask . 3. An illumination optical system for irradiating an exposure illumination light onto a mask, a projection optical system corrected for aberrations with respect to the exposure illumination light, and a responsive mask pattern being transferred via the projection optical system. A projection exposure apparatus comprising: a stage for holding a substrate; and an alignment mark on the sensitive substrate, or a first reference mark disposed on the stage, a detection illumination light having a wavelength different from the exposure illumination light. Irradiate the alignment
And a first mark detection system for detecting optical information generated from the first reference mark, and detection of the first reference mark by the first mark detection system.
At a position different from the position,
The exposure illumination is applied to a second reference mark different from the first reference mark.
Irradiate light, and the mask on the mask is projected through the projection optical system .
A second mark detection system for detecting relative position information with respect to the mark and the first and second reference marks move the stage.
Detected by the first and second mark detection systems without any
A projection exposure apparatus wherein the positional relationship is determined so as to be performed. 4. An adjustment device for adjusting optical characteristics of the projection optical system, wherein the first and second mark detection systems detect the first and second reference marks after adjusting the optical characteristics.
The projection exposure apparatus according to any one of claims 1 to 3, characterized in Rukoto issue. 5. A projection exposure method for irradiating a mask with exposure illumination light and transferring a pattern of the mask onto a sensitive substrate via a projection optical system, wherein a reference plate having first and second reference marks formed thereon. Is disposed on the sensitive substrate side with respect to the projection optical system , and the first mark detection using detection illumination light having a different wavelength from the exposure illumination light.
Simultaneously detecting the first reference mark by out system,
Detection of the first reference mark by the first mark detection system
In position a different position, the detecting and the second reference mark and the mark on the mask by the second mark detecting system using the exposure illumination light, the first and second mark detection system
Alignment between the mask and the sensitive substrate
A projection exposure method characterized by being used in a projection exposure method. 6. The reference plate according to claim 5, wherein the reference plate is substantially stationary while the first and second mark detection systems detect the first and second reference marks, respectively. Projection exposure method. 7. The optical characteristic of the projection optical system, or the first
After adjusting the position of the mark detection system, the first and second reference
7. The projection exposure method according to claim 5, wherein a mark is detected . 8. A projection exposure method of irradiating a mask with exposure illumination light and transferring a pattern of the mask onto a sensitive substrate via a projection optical system, wherein the detection illumination light has a wavelength different from that of the exposure illumination light. the light information generated from the first reference mark and detects the first mark detection system by irradiation of, in the first mark detection system
That at the detection position and different positions of the first reference mark, the projection optical system and the second reference mark different from the first reference mark through the mark on the mask, second using the exposure illumination light The first reference mark is detected by a mark detection system .
Between the detection of the reference mark and the detection of the second fiducial mark.
Disposed on the sensitive substrate side with respect to the projection optical system, and
A movable body on which the first and second reference marks are integrally provided
A projection exposure method, which is performed without moving the image . 9. An illumination optical system for irradiating a mask with exposure illumination light.
System and projecting a pattern image of the mask on a sensitive substrate
A projection optical system, and a first detecting a mark on the sensitive substrate.
A projection exposure apparatus provided with a mark detection system, the projection exposure apparatus having a mark detection system different from the first mark detection system within the field of view of the projection optical system
At the center of the projection optical system and located on the image plane side of the projection optical system.
The mark to be placed is detected together with the mark on the mask.
A second mark detection system disposed on the image plane side of the projection optical system;
Detection of the first fiducial mark by the output system and detection of the second mark
The detection of the second fiducial mark by the outgoing system can be executed simultaneously.
The first and second fiducial marks are provided integrally in a positional relationship.
A projection exposure apparatus, comprising: 10. The first and second reference marks are movable.
The shape is determined by the positional relationship where the detection is performed without moving the body.
10. The projection exposure apparatus according to claim 9, wherein the projection exposure apparatus is formed.
Place.