JPH01164031A - Alignment device - Google Patents

Abstract

PURPOSE:To make highly accurate alignment by scanning spots produced by irradiation light for alignment or special patterns to obtain relative positional information between a mask and an exposure substrate. CONSTITUTION:When the angle of a vibration mirror 26 changes continuously, incidence direction of an irradiation light LB to a reticule mark changes continu ously. When positions of spots SU and SD change, positions of spot images formed at the position in cooperation with these change. Spot positions on slits 44, 50, 56, and 58 are measured by a grating 42. When a spot pattern PG caused by incidence light moves in the direction of an arrow F1 as a vibration mirror 26 vibrates, the intensity of penetrated light changes in the form of a sine wave or triangle wave. This change is detected by a detector 46. Processing, for example, signals of frequency division, etc., obtains position pulses. By counting this, spot positions can be measured.

Description

[Detailed Description of the Invention] [Field of Application in Industry A] The present invention relates to an alignment device for, for example, a projection exposure apparatus, and particularly to an alignment device that uses light of a wavelength different from exposure light as alignment light. This is related to the improvement of

[Prior Art] There are two types of alignment devices for exposure apparatuses: one uses light of the same wavelength as the exposure light and the other uses light of a different wavelength as illumination light for alignment.

Among these, an example of a conventional alignment apparatus that uses light of a wavelength different from that of exposure light is shown in FIG.

In this figure, a predetermined circuit pattern formed on a mask or reticle R is projected onto a wafer W on a wafer stage 12 via a projection lens 10 by exposure light indicated by dotted lines.

When exposing such a circuit pattern, it is necessary to align the reticle R and the wafer W as described above.

The alignment light, which has a non-exposure wavelength, is irradiated from the side from a light source (not shown) to the mirror 14, as shown by the dashed line in the figure, and is reflected there and enters the reticle R.

Further, the alignment light that has passed through the reticle R passes through the correction optical system 16 and enters the projection lens 10, and then passes through this and enters the wafer W. Furthermore, the alignment light reflected by the wafer W returns on its outward path and enters a detector (not shown).

Note that the correction optical system 16 corrects chromatic aberration caused in the projection lens 10 with respect to alignment light or imaging light.

The alignment mark formed on the reticle R and the alignment mark formed on the wafer W are each detected under the irradiation of alignment light by the action of the correction optical system 16, and the positional relationship between the two is detected by the output of the detector. The positional deviation between the reticle R and the wafer W is detected from the signal. By moving the wafer stage 12 based on this, alignment between the reticle R and the wafer W can be performed.

After the alignment is performed as described above and the reticle R and wafer W are set at the alignment position, the mirror 14 and the correction optical system 16 are moved out of the exposure light optical path, and exposure is performed.

On the other hand, in a method that uses light of the same wavelength as the exposure light, no chromatic aberration occurs in the projection lens 10 with respect to the alignment light or the imaging light, so the above-mentioned correction optical system is not necessary.

[Problem to be Solved by the Invention 1 Among the conventional techniques described above, first, in the method using light of a wavelength different from the exposure wavelength, a part of the alignment optical system is retracted when performing exposure. ing.

For this reason, the configuration becomes extremely complicated, and as the evacuation operation is repeated, the positioning of the optical elements may become misaligned.
As a result, alignment accuracy may be reduced. Further, due to the structure, it is impossible to perform alignment during exposure.

Furthermore, if the flatness of the glass substrate surface constituting the mask (or reticle) is not good, or if a taper occurs due to deflection due to the mask's own weight, it is necessary to irradiate the alignment light (reticle mark light image) onto the wafer. Misalignment may occur. Alignment accuracy also decreases due to such causes.

Next, with methods that use light with the same wavelength as the exposure wavelength or light with a wavelength close to it, alignment accuracy may decrease or alignment may become impossible depending on the type of resist on the wafer W. be.

For example, if there is aluminum or the like in the base, a resist that absorbs the exposure light may be used to prevent poor exposure due to reflection of the exposure light.

In such a case, the alignment light will be absorbed by the resist, making it impossible to detect a sufficient amount of returned light to detect the alignment mark, resulting in decreased alignment accuracy or, in some cases, alignment failure. becomes.

The present invention has been made in view of the above, and provides an alignment device that can perform good alignment with high precision even though light with a wavelength different from the exposure wavelength is used as illumination light for alignment. Its purpose is to

[Means for Solving the Problems] The present invention has a unique focal length corresponding to the amount of chromatic aberration of the projection optical system caused by the difference between the sensitive wavelength of the photosensitive substrate and the wavelength of the illumination light, and the irradiation of the illumination light. forming a special pattern on the mask that has lens characteristics to form a light spot of a predetermined shape in space; and scanning means for continuously changing the angle of incidence of illumination light with respect to the special pattern; and a light spot formed by the special pattern. a first photoelectric detection means for receiving detection light generated from the spot when a mark formed on the photosensitive substrate is scanned; a first photoelectric detection means for receiving the light spot or its conjugate image; a twentieth photoelectric detection means for outputting a signal representing the position of the light spot or its conjugate image that changes due to the action; 3. outputting a signal representative of the position with respect to the vertical illumination state of the special pattern and the vertical projection state of the conjugate image of said light spot; a fourth photoelectric detection means; the output signals of the third and tJ4 photoelectric detection means are detected in correspondence with the output signal of the second photoelectric detection means, and the alignment between the mask and the substrate is determined based on the detection result; Offset detection means for measuring an offset amount: an output signal of the first photoelectric detection means is
Technically, the invention further comprises a position detection means that detects the output signal in correspondence with the output signal of the photoelectric detection means and obtains relative position information between the mask and the photosensitive substrate based on the detection result and the offset amount. This is the main point.

[Function] According to the present invention, a special pattern with a unique focal length having a lens effect is formed on the mask. When coherent or quasi-monochromatic illumination light is incident on this special pattern, a light spot is formed at a spatial position corresponding to the amount of chromatic aberration of the projection optical system, and the chromatic aberration of the projection optical system with respect to the illumination light is corrected.

The angle of incidence of the illumination light on the special pattern is continuously changed by the scanning means. Due to this action, the spot light scans the mark formed on the photosensitive substrate, and the detection light is incident on the first photoelectric detection means.

On the other hand, the spot light on the mask side resulting from such scanning is incident on the second photoelectric detection means for position detection.

Further, the conjugate image of the spot light is received by third and fourth photoelectric detection means on the illumination light incident side of the mask and the photosensitive substrate side, respectively, and the conjugate image of the light spot is detected in the vertical illumination state of the special pattern and the conjugate image of the light spot. Signals representing positions with respect to vertical projection conditions are each detected.

The output signals of the third and fourth photoelectric detection means are outputted in correspondence with the output signals of the second photoelectric detection means and inputted to the offset detection means, and the amount of alignment offset between the mask and the substrate is measured. be done.

Next, the output signal of the first photoelectric detection means is detected in correspondence with the output signal of the second photoelectric detection means, and is input to the position detection means together with the offset amount, so that the relative relationship between the mask and the photosensitive substrate is detected. Location information can be obtained.

[Embodiments] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. Note that the same reference numerals are used for the same parts as in the prior art described above.

FIG. 1 shows the optical components of one embodiment of the invention. In this figure, a dichroic mirror 20 is arranged on the exposure light incident side of the reticle R.
Exposure light LA is made to enter from the right side of the figure, as shown by the dotted arrow. The exposure light LA is reflected by the dichroic mirror 20, passes through the reticle R, and enters the projection lens 22.

The exposure light LA is caused by the action of the projection lens 22 to be incident on the wafer W set on the wafer stage 24, so that the circuit pattern of the reticle R is projected thereon. The wafer stage 24 described above is
It is configured to be able to move the wafer W in the Y and Z directions.

Next, illumination light LB for alignment having a different wavelength from the exposure light LA is made to enter the vibrating mirror 26 as parallel light from a light source (not shown) as indicated by a dashed-dotted line arrow. The illumination light LB reflected by the vibrating mirror 26 further passes through the half mirror 28 , is reflected by the half mirror 30 via the lens system 29 , and is reflected by the dichroic mirror 20 .
is configured to be incident on the The origin of the deflection of the illumination light LB of the vibrating mirror 26 is determined at the rear focal position of the lens system 29.

Next, the illumination light LB transmitted through the dichroic mirror 20 enters the reticle R, passes through each of the reticle R1 projection lenses 22, and enters the wafer W. Here, the reticle R and the vibrating mirror 26 are determined to be conjugate with each other.

Optical information reflected and scattered on the surface of the wafer W (hereinafter referred to as
The LC (referred to as "r detection light") returns in the opposite direction along the optical path of the illumination light LB described above and enters the dichroic mirror 20 again.

The light passes through this dichroic mirror 20 and passes through the half mirror 30 and lens system 29 to the half mirror 28.
The detection light LC reflected by each of the detectors is condensed by a condenser lens 32, and is made to enter a detector 34 that performs photoelectric conversion. This detector 34 is located at a position conjugate with the pupil plane of the projection lens 22.

Next, the above-mentioned reticle R has a configuration as shown in FIG. 3, for example. In this figure, a circuit pattern to be projected onto a wafer W is formed in the center portion of the reticle R, and the edges excluding the pattern area RP are provided with a circuit pattern corresponding to the X1Y directions in FIG. Reticle marks RMX and RMY for alignment are formed respectively. The centers of these reticle marks RMX, RMY correspond to the center RQ of the reticle R as shown by the dashed line.

FIG. 4 shows an example of a pattern of reticle marks RMX, RMY. As shown in this figure, the pattern shape is a linear Fresnel pattern, and its center is
It corresponds to the center RQ of the reticle R. For this reason, Mark RMX.

When parallel illumination light (laser light) LB enters RMY,
A slit-shaped beam focused in the -dimensional direction is imaged. This is because a linear Fresnel pattern functions equivalently to a cylindrical lens.

Therefore, the reticle mark RMX as shown by the arrow FA in Figure 3.
, RMY, due to the lens effect of the Fresnel pattern, a spot Sυ, SD ( The beam is imaged here in the form of a slit). and,
From each of these spots SU and SD, the illumination light flux spreads isometrically and advances. A spot SWU, which is an image of the spot SU conjugately formed by the projection lens 22, is formed on the surface of the wafer W described above, and is located at a position conjugate with the spot SU (see FIG. 1).

Next, the light from the Fresnel pattern that has passed through the half mirror 30 described above is focused by the condenser lens 36 and then enters the half mirror 38, and the light that has passed through this is
The beam is configured to be incident on the half mirror 40.

Among these, on the optical axis of the light reflected by the half mirrors 38 and 40, a grating 42 and a slit 44 are arranged at positions conjugate with the spot SU, and the transmitted light is subjected to photoelectric conversion. Detector 46.4 for
8, respectively.

On the other hand, a slit 50 is arranged on the optical axis of the light transmitted through the half mirror 40 at a position conjugate with the spot SD, and is configured so that this transmitted light is input to a detector 52 for photoelectric conversion. ing.

Next, a detection section 54 as shown in FIG. 1 is provided on the wafer stage 24 described above. This detection section 54
As shown enlarged in FIG. 6(A), has slits 56 and 58 formed with a predetermined step, and the light that passes through these as indicated by the arrow FB%FC is The light is incident on detectors 60 and 62 for photoelectric conversion, and the optical signals are converted into electrical signals.

Among these slits 56 and 58, the slit 56 is
The position conjugate to the spot SU in FIG. 5, that is, the wafer W
It is arranged and set at the position of the spot SWU (see FIG. 1) formed on the surface of. Further, the slit 58 is located at a position conjugate with the spot SD in FIG.
WD (see FIG. 1).

When the angle of the vibrating mirror 26 mentioned above changes continuously,
The direction of incidence of the illumination light LB on the reticle marks, RMX, and RMY changes continuously. That is, when the vibrating mirror 26 vibrates, the direction of incidence of the illumination light LB changes continuously as indicated by the arrow FD in FIG. 3 or FIG. 5. Therefore, the positions of the spots SU%SD also change continuously in opposite directions.

When the positions of the spots SLI and SD change in this way, the position of the spot image formed at a position conjugate to them also changes. The spots SWU and SWD mentioned above are
The position changes as shown in FIGS. 6(B) and 6(C), respectively. That is, when spot SWU moves in the direction of arrow FE, spot SWD moves in the direction of arrow FF.

Also, when spot SWU moves in the direction of arrow FG,
Spot SWD moves in the direction of arrow FH.

The position of the spot on such a slit 44, 50, 56, 58 is adapted to be measured by the grating 42.

FIG. 7(A) shows an example of a badan of the grating 42, which has a configuration in which rectangular patterns are continuously arranged. The spot pattern PG due to the incident light is
When the vibrating mirror 26 moves in the direction of the arrow FI as it vibrates, the intensity of the transmitted light changes in the form of a sine wave or a triangular wave, as shown in FIG. When this detection signal is subjected to signal processing such as frequency division, a position pulse as shown in FIG. 2(C) can be obtained. By counting this, the spot position (scanning position) can be measured.

Next, the above-mentioned wafer W has a plurality of shot areas E on which a predetermined pattern corresponding to a unit chip is projected, as shown in FIG. 8(A), for example, and the pattern area RP of the reticle R is The circuit pattern is scaled down and projected across each shot area.

At the edge of each shot area E, wafer marks EMX and EMY for alignment are formed respectively corresponding to the X and Y directions in FIG. 1, as shown in FIG. 1B. The centers of these wafer marks EMX and EMY correspond to the center EQ of the shot area E, as shown by the dashed line.

FIG. 9 shows an example of wafer marks EMX and EMY, which are formed on the wafer W as surface irregularities. The slit-shaped spot light irradiated onto the wafer W as indicated by the arrow FJ is caused by the vibration of the vibrating mirror 26.
Move the top in the direction of arrow FK and select the wafer mark EMX%EM
It starts scanning Y. Note that since the scanning position of this slit-shaped spot light corresponds to the angle of the vibrating mirror 26, it also corresponds to the spot position detected by the grating 42 shown in FIG. 7.

Next, the signal processing system of each detector will be explained with reference to FIG.

In Figure 10, detector 34.48.52.60
．． The output sides of 62 are connected to the input sides of memories (RAM) 68A to 68E, respectively, via amplifiers 64A to 64E and ADCs 66A to 66E, which convert analog signals to digital signals.

Furthermore, the output side of the RAMs 68A to 68E is the waveform processing section 7.
0, and the processing output side of this waveform processing section 70 is connected to the drive control section 72 of the wafer stage 24 shown in FIG. A position detector 74 for detecting the coordinate position of the wafer stage 24 using an interferometer or the like is connected to the drive control unit 72, and a drive motor 76 for driving the wafer stage 24 is also connected. ing.

On the other hand, the other detector 46 is an amplifier 78? The pulsing circuit 80 is connected to the pulsing circuit 80 via the pulsing circuit 80.
On the output side of 0, in addition to the counter 82, the above-mentioned ADC 6
6A to 66E, respectively, so that sampling pulses are output. Further, the count output of the counter 82 is connected to the RAMs 68A to 68E so as to be input as address data.

, Among the above units, the waveform processing unit 72 has a RAM 66.8.
Each detector 34.48.5 stored in 8.70 respectively
Based on each of the detection signals No. 2, the amount of deviation between the reticle R and the wafer W, that is, the alignment error is calculated.

Next, the drive control unit 70 refers to the position of the wafer stage 24 determined by the position detector 74, controls the drive controller 76, and moves the stage by the amount of deviation inputted. .

Further, the pulsing circuit 80 performs pulsing as shown in FIG. 7 on the input signal.

Next, the overall operation of the above embodiment will be explained with reference to FIGS. 11 and 12 in addition to the above-mentioned drawings.

First, for ease of understanding, assume that the illumination light LB is incident on the reticle R perpendicularly to the reticle R, that is, in the same way as the exposure light LA, and that the reticle R is not tapered. explain.

As described above, the illumination light LB has a different wavelength from the exposure light LA. Since the projection lens 22 has chromatic aberration corrected for the exposure light LA, a certain chromatic aberration occurs for the illumination light LB which is a non-exposure wavelength.

Therefore, for light having the exposure wavelength, the light output from the spot SA at the position of the reticle R forms an image on the wafer W via the projection lens 22, as shown in FIG. 11(^).

However, for light at non-exposure wavelengths, for example,
As shown in FIG. 2, the light output from the spot SB of the reticle R is imaged Δ below the wafer W. This Δ
is the amount of chromatic aberration.

In such a case, in order to form an image on the wafer W,
As shown in FIG. 4C, light may be output from a spot SC formed at a position Δ' above the reticle R.

Therefore, in this embodiment, reticle marks RMX, RM
As Y, a Fresnel pattern as shown in Fig. 4 is used, and due to the lens effect caused by this, Fig. 11 (C)
The effects shown are obtained.

That is, the position of the spot SU in FIG. 5 and the position of the surface of the wafer W are configured to be conjugate with respect to the projection lens 22.

Note that the magnitude of Δ (or Δ') can be adjusted by changing the Fresnel pattern shown in FIG. Furthermore, when the amount of chromatic aberration on the wafer W side of the projection lens is Δ, and the reduction ratio of the projection lens is 17M, the amount of deviation Δ' of the spot SU from the reticle required to correct this Δ is Δ' w M It is defined as 2・Δ.

As described above, the projection lens 22 for the illumination light LB
The chromatic aberration is corrected, and the illumination light LB is imaged on the wafer W as a spot SWU.

Next, when the illumination light LB is perpendicularly incident on the reticle marks RMX, RMY, it is the same as that the reticle R is exposed by the exposure light.

That is, as described above, the reticle marks RMX, RM
Since the center of Y corresponds to the center RQ of reticle R,
The centers of the images of the reticle marks RMX and RMY projected onto the wafer W by the illumination light LB of the vertical illumination correspond to the center RQ of the reticle R.

On the other hand, wafer mark EMX in shot area E.

The center of EMY also corresponds to the center EQ of shot area E.

Therefore, in this state, the difference DA (see FIG. 8) between the imaging position of the slit-shaped spot light based on the illumination light LB on the wafer W and the center of the wafer mark EMX%EMY (see FIG. 8) indicates that the positions of the two do not deviate. This will result in real alignment error.

Therefore, in order to perform alignment, the above-mentioned vertical illumination state, specifically, the reticle marks RMX, RMY
It is necessary to find the angle of the vibrating mirror 26 or the illumination position of the illumination light LB in a state where the illumination light LB is incident perpendicularly to the oscillating mirror 26.

In this embodiment, the vertical illumination position of the illumination light LB is determined from the output of the detector 48.52. The operation of determining this vertical illumination position will be described below.

First, by setting the vibrating mirror 26 at an appropriate angle, the illumination light LB
Reticle R's reticle mark RMX.

Irradiate RMY.

As shown in FIG. 5, the illumination light LB irradiates the linear Fresnel pattern (marks RMX, RMY) to form spots SU above and below the tickle R.

Form SD. And these spots SU, SD
The imaging light generated from the slits 44 and 50 forms a slit-like image on the surface of each slit 44 and 50, and is further transmitted through the slits 44 and 50 to the detector 4.
8.52 respectively. Note that an image of the spot light is also formed on the surface of the grating 42, and the image ray is also incident on the detector 46.

In this state, when the vibrating mirror 26 is vibrated,
As shown in FIG. 5, 5f of illumination light LB in the direction of arrow FB
Along with JjJ, there's also sports.

The grating 42 and the slits 44.5
The position of the spot on 0 will also move.

In the above operation, the illumination light LB is
In the state where the incident is perpendicular to X, RMY, the detector 48
．． The amount of light of the spot image incident on 52 becomes the largest,
The output waveforms of the detectors 48 and 52 are as shown in (A) in FIG.
The results are shown in (B).

In these figures, the peak position RA of the detector 48
and the average position (R
By determining A+RB)/2, this can be used as the vertical illumination position of the illumination light. Hereinafter, this vertical illumination position will be expressed as α.

(RA+RB)/2=α・・・・・・・・・・・・・・・
(1) The above explanation is for the case where the reticle R has no taper.

Next, measurement of a change in the spot position on the wafer W due to the occurrence of surface taper due to the self-Ii of the reticle R will be explained.

If the flatness of the surface of the glass substrate constituting the reticle R is not good or if a certain taber exists due to its own weight, the illumination light L
Even if B enters and passes through, it does not enter the wafer W perpendicularly through the projection lens 22, causing a shift in the irradiation position of the slit-shaped spot light. Therefore, it is necessary to determine the amount of such deviation and perform correction at the time of alignment.

The above-mentioned change in the position of the slit-shaped spot light on the wafer W (particularly the change in the measurement direction perpendicular to the longitudinal direction) is similar to the method of determining the vertical incident position of the illumination light beam B on the reticle R described above. be.

That is, by moving the wafer stage 24, the detection unit 5.
In the vertical incidence state when the slit-shaped spot light imaged by the projection lens 22 is made to enter 4,
Since the slit-shaped spot light and each slit 56.58 are parallel to each other, the amount of light transmitted through the slit 56.58 and incident on the detector 60.62 is maximized.

Therefore, the output waveforms of the detectors 60 and 62 when the vibrating mirror 26 is vibrated are as shown in FIGS. 12(C) and 12(D), respectively, with respect to the measurement position based on the output of the detector 46.

In these figures, the peak position RC of the detector 60
and the peak position RD of the detecting pattern 62 (
RC+RD)/2, this is the slit-shaped spot light SWU.

It can also be a vertical illumination position of the SWD. Hereinafter, this vertical illumination position will be expressed as β.

(RC+RD)/2-β・・・・・・・・・・・・・
(2) If the alignment is performed in consideration of α and β obtained as described above, the deviation amount DA shown in FIG. 9 can be favorably corrected. In other words, α−β=γ・・・・・・・・・・・・・・・・・・
Taking into account (3), the alignment position of the shot area E of the wafer W with respect to the reticle R is determined.

Next, the overall operation of the above device will be explained.

First, the detection unit 54 receives the slit-shaped spot lights SWU and SW.
When the wafer stage 24 is moved and the vibrating mirror 26 is vibrated so that the beam D is incident, the illumination light LB incident on the reticle mark RMX%RMY of the reticle R is continuously tilted as shown in FIG. As a result, the positions of the spots SU%SD formed on the top and bottom of the reticle R change, and the spot positions on the grating 42, the slits 44, 50, and the detection section 54 change, respectively.

These spot position changes are detected by detectors 48, 52.
46, 60, and 62, respectively, and the amplifiers 64B to 64
The signals are input to E and 78, respectively, and signal amplification is performed.

The output of the amplifier 78 is input to a pulsing circuit 80,
In addition to the counter 82, the pulse indicating the position is sent to the ADC 66B.
~66E are each input as sample pulses.

The ADCs 66B to 68E digitize the input signals in synchronization with the timing of the input sample pulses, and the signal waveforms shown in FIG. 12 are stored as digital signals in the RAMs 68B to 68H, respectively. The address of this paddy is specified by the count value of the counter 82. That is, the count value of the counter 82 is position data, and the waveform data corresponding to this is stored at the corresponding address.

Next, in the waveform processing unit 70, the waveform data stored in the RAMs 68B to 88H are read out, and first, the vertical illumination positions α and β of the illumination light LB and the slit-shaped spot lights SWU and SWD and the offset amount γ are are calculated respectively.

On the other hand, move the wafer stage 24 to mark the wafer mark EM.
The slit-shaped spot light SWU is incident on X and EMY (at this time, the luminous flux of the spot light SWD is largely defocused on the square plane), and when the vibrating mirror 26 is vibrated as described above, the wafer W wafer mark E
MX.

Detection light from EMY enters the detector 34. The output waveform of this detector 34 is, for example, (E) in FIG.
It becomes as shown in . If the peak position of this waveform is RE, then (RA+RB)/2+γ-RE...
- (4) represents the amount of positional deviation between the reticle R and the shot area E of the wafer W.

As described above, the amount of positional deviation DA between the reticle R and the shot area E of the wafer W is calculated by the waveform processing unit 70.

This positional deviation amount is input to the drive control section 72. In the drive control unit 72, based on this input, the position detector 74
A drive command is issued to the drive motor 76 while taking into account the coordinate position of the wafer stage 24 monitored by the wafer stage 24, the positional deviation 3iDA is corrected, and the alignment is completed.

Next, the circuit pattern of reticle R is exposed.
Since the exposure light LA is incident on the reticle R via the dichroic mirror 20, there is no need to evacuate the optical element as in the prior art shown in FIG.

As explained above, according to this embodiment, the exposure light LA
Since the exposure light and the alignment illumination light are separated by disposing a dichroic mirror that reflects the illumination light LB and transmits the illumination light LB, alignment can be continued even during exposure.

Furthermore, in order to correct the chromatic aberration of the illumination light for alignment with respect to the projection optical system, a Fresnel pattern is placed on the reticle R, so that a correction optical element or the like is not required, and the configuration is simplified.

Furthermore, since alignment is performed using light at non-exposure wavelengths,
Even when a special resist or the like that absorbs exposure light is used, there is an effect that alignment can be performed satisfactorily without any inconvenience.

In addition, detection of spots formed above and below the reticle R,
When detecting the detection light from the wafer W, such a spot is used to generate a sampling pulse for scanning position detection, so that the influence of mechanical or optical fluctuations of the device on the measurement is reduced. Highly accurate alignment can be performed.

Note that the present invention is not limited to the above-mentioned embodiments; for example, in the above-mentioned embodiments, scanning may be performed by moving the wafer stage 24 instead of scanning the illumination light LB by the vibration of the mirror 26. Specifically, the illumination light LB is adjusted to be perpendicularly incident on the reticle R, and the wafer mark is moved under the illumination spot.

This method of moving the stage has an advantage over the method of using a vibrating mirror in that it does not require a vibrating mirror and its driving part, but in order to move the stage,
This is disadvantageous in that alignment cannot be performed during exposure.

In addition, the scanning position of the spot light for alignment is
The measurement may be performed using the means shown in FIG. This example is effective when, in the embodiment shown in FIG. 1, the diameter of the spot created by the illumination light LB and the Fresnel pattern on the grating 42 is large and it is difficult to obtain a good position pulse.

In FIG. 13, the vibrating mirror 100 has mirror surfaces on the front and back, and the illumination light LB incident from the top of the figure is reflected to the right side of the figure, and the illumination light LB incident from the left side of the figure is used for scanning position detection. The light is reflected downward in the figure and is configured to enter a grating 104 via a condenser lens 102. As the vibrating mirror 100 vibrates, the grating 1
The detection light spot on 04 also moves, and a pulse signal for scanning position detection can be obtained satisfactorily.

Alternatively, sample pulses for scanning position detection may be generated at regular time intervals.

In addition to the Fresnel pattern, the reticle mark may be of any type as long as it has a lens effect.

Furthermore, in the above embodiment, in order to detect the offset amount, the movement state of the alignment light spots SU and SD formed above and below the reticle R is detected, but basically, the movement state of the spots SU and SD of the alignment light formed above and below the reticle R is detected. It is sufficient to detect movement of either one.

[Effects of the Invention] As explained above, according to the present invention, a pattern having a lens effect is formed on a mask, so that when light of a wavelength different from the exposure wavelength is used as illumination light for alignment, In addition to correcting the chromatic aberration that occurs in the projection optical system, we decided to scan a spot created by the alignment illumination light or a special pattern to obtain relative positional information between the mask and the photosensitive substrate, so we created a special correction optical system. This has the advantage that good alignment with high accuracy can be performed without the need for. Another advantage is that alignment can be continued even during exposure.

[Brief explanation of the drawing]

FIG. 15 is a configuration diagram showing the optical components of this invention,
FIG. 2 is a configuration diagram showing an example of a conventional device, FIG. 3 is a perspective view showing an example of a reticle and how alignment light is incident, and FIG.
The figure is a plan view showing an example of a pattern of a reticle mark, FIG. 5 is an explanatory diagram showing the vibration of alignment light due to vibration of a vibrating mirror, FIG. 7 is a diagram showing the operation of position detection using pulses, FIG. 8 is a perspective view showing an example of a wafer, FIG. 9 is a partial perspective view showing an example of a wafer mark, and FIG. 10 is an electrical diagram of the above embodiment. FIG. 11 is a block diagram showing the constituent parts, FIG. 11 is an explanatory diagram of chromatic aberration correction of the projection lens, FIG. 12 is a diagram showing the action during alignment, and FIG. 13 is an explanatory diagram showing other position detection means. 20... Dichroic mirror, 22... Throwing lens, 24... Wafer stage, 26... Vibration mirror, 34.46, 48.52, 60.62... Detector, 42... Gray Proboscis, 44, 50, 58° 58
...Slit, 70...Waveform IA science department, 72...
Drive control unit, LA...exposure light, LB...alignment light, LC...detection light, R...reticle, RMX,
RMY...Reticle mark, SU, SD...Spot, W...Wafer, EMX, EMY...Wafer mark.

Claims (1)

[Scope of Claims] An apparatus for aligning a mask and a photosensitive substrate via a projection optical system using illumination light having a wavelength different from the wavelength at which the photosensitive substrate is sensitive, comprising: forming on the mask a special pattern having a unique focal length corresponding to the amount of chromatic aberration of the projection optical system that occurs, and having lens characteristics that form a light spot of a predetermined shape in space by irradiation with the illumination light; , a scanning means for continuously changing the incident angle of illumination light with respect to the special pattern of the mask, and detection generated from the mark formed on the photosensitive substrate when a light spot formed by the special pattern scans the mark. a first photoelectric detection means that receives light; and a second photoelectric detection means that receives the light spot or its conjugate image and outputs a signal representing the scanning position of the light spot or its conjugate image that changes due to the action of the scanning means. a photoelectric detection means; receiving a conjugate image of the light spot on the illumination light incident side of the mask, and outputting a signal representing the position of the conjugate image of the light spot when the special pattern is vertically illuminated with illumination light; a third photoelectric detection means; a fourth photoelectric detection means for receiving a conjugate image of the light spot on the photosensitive substrate side and outputting a signal representing a vertical projection position of the conjugate image of the light spot; and offset detection means for detecting the output signal of the fourth photoelectric detection means in correspondence with the output signal of the second photoelectric detection means, and measuring the amount of alignment offset between the mask and the substrate from the detection result; The output signal of the first photoelectric detection means is detected in correspondence with the output signal of the second photoelectric detection means, and the relative position between the mask and the photosensitive substrate is determined based on the detection result and the offset amount. An alignment device comprising: position detection means for obtaining information.