JP2541219B2 - Projection optical device - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a projection optical apparatus, and more particularly to an improvement in a method for measuring distortion or magnification of a projection optical system.

[Prior Art] Conventionally, as a method of measuring the distortion of a projection optical system used in an optical apparatus, in particular, an exposure apparatus, a method of printing a mark on a semiconductor wafer has been performed as shown in FIG. 8, for example. .

This method is a method of converting the distortion amount of the exposure projection optical system into a short dimension by using an interferometer that measures the coordinate position of the wafer stage.

In FIG. 8, first, on the reticle R, patterns P1 to P5 for distortion measurement are formed respectively (see FIG. 8A).

Using the reticle R as described above,
The patterns P1 to P5 are printed by whole surface exposure (see FIG. 3B).

Next, of the marks on the reticle R, the center mark
A blind is applied except for P3, and the wafer W is exposed again (see FIG. 7C).

At this time, the coordinate value of the stage is set based on the design data known in advance, and the mark P3 is overprinted on the already burned marks PA or PE (see (D) in the same figure).

As described above, the marks formed by overlapping are exactly overlapped if no distortion exists in the projection optical system. However, since there is distortion, there is a gap Δ between the marks as shown in FIG. Occurs. The amount of distortion of the projection optical system can be measured by this Δ.

[Problems to be Solved by the Invention] However, in the method described above, exposure of the wafer,
It requires a process in which the development is repeated twice, followed by measurement. Therefore, there is an inconvenience that the measurement is troublesome.

In addition, since error factors such as stage stop accuracy and pattern measurement accuracy directly affect distortion measurement accuracy, there is the inconvenience that the measurement reproducibility is not good.

The present invention has been made in view of the above points, and an object thereof is to provide a projection optical apparatus capable of accurately measuring distortion and magnification in a short time.

[Means for Solving Problems] According to the present invention, a stage that can move along an image plane of a projection optical system that projects a pattern on a mask; and a position signal that corresponds to the position of the stage is output. Stage position detecting means; light emitting means provided on the stage and having a light emitting surface of a predetermined shape; photoelectric detecting means for receiving light from the light emitting surface of the light emitting means through the projection optical system and the mask, respectively. Controlling the stage so that a projected image of the light emitting surface moves with respect to a mark pattern formed at a predetermined position of the mask, and a photoelectric signal output from the photoelectric detection means during the movement; Position detecting means for detecting a positional relationship between the projected image of the light emitting surface and the max pattern based on a position signal output from the stage position detecting means; The numerical aperture of the surface for the mask is set equal to the numerical aperture of the illumination light at the time of pattern projection for the mask; the technical point is that the σ value of the measuring illumination by the light emitting means is set to 1 or more. is there.

[Operation] According to the present invention, first, the numerical aperture for the mask of the light receiving surface of the photoelectric detection unit that receives the light from the light emitting unit is set equal to the numerical aperture for the mask of the illumination light at the time of pattern projection.

In the above state, the stage is moved, that is, the light emitting means is moved, and the position detecting means detects the positional relationship between the projected image of the light emitting means and the mask pattern.

The data obtained by these operations and the design data are referred to, and the imaging characteristics of the projection optical system are obtained.

This measured characteristic matches the characteristic of the projection optical system when the mask pattern is projected.

EXAMPLES Examples of the present invention will be described in detail below with reference to the accompanying drawings.

Overall Configuration of Embodiment FIG. 1 shows the overall configuration of an embodiment of the present invention. In this figure, illumination light for exposure (eg g line, i line) output from an appropriate light source for exposure (not shown)
LA is incident on a dichroic mirror 16 via a fly-eye lens 10, a low-reflectance beam splitter 12, and a relay optical system 14, respectively.

The illumination light LA reflected by the dichroic mirror 16 passes through the condenser lens 18 and enters the reticle R. Further, the illumination light LA passes through the projection lens 20 having a telecentric characteristic on the image side and the wafer W on the stage 22. It is designed to illuminate.

In the vicinity of the reticle R described above, a TTL alignment microscope 26 for aligning the reticle R set on the reticle holder 24 and an alignment between the reticle R and the wafer W, and a TTL alignment microscope 28 dedicated for wafer alignment. Are arranged respectively.

Further, a focus sensor 30 for detecting a focus position of a wafer or a slit plate described later, and an off-axis microscope for wafer alignment are provided on the side of the projection lens 20.
32 and 32 are arranged respectively.

Next, the above-mentioned stage 22 is provided with a slit plate 34 at the surface position of the wafer W.

FIG. 2 shows an enlarged cross section of the slit plate 34. In this figure, a stage 22 is a Z stage 22A that is movable in the Z direction, which is the vertical direction in the figure.
And an XY stage 22B that is movable in the XY direction that is a direction perpendicular to the Z direction, and a θ table 22C that can be minutely rotated. The wafer W is placed on the θ stage 22C, and the slit plate 34 is held on the Z stage 2 by the holder 34A.
Placed and held on 2A.

In addition, inside the Z stage 22A, a lens 34B and a mirror 34
C and a fiber 34D are respectively arranged, and the light LB of the light source 34E (see FIG. 1) that outputs the light having the same wavelength as the exposure wavelength provided outside illuminates the slit plate 34 from the inside of the stage. Has become.

Here, the slit plate 34 of the optical LB that proves the slit plate 34
The characteristics and arrangement of each optical element are set such that the numerical aperture for the projection lens 20 is larger than the numerical aperture for the wafer W of the projection lens 20.

Next, on the side of the Z stage 22A, a movable mirror 36 for a laser interferometer for measuring the coordinate position of the stage is provided, and as shown by an arrow FA, the height position of the slit plate 34 and The laser beam is irradiated at almost the same height.

Next, the slit plate 34 will be described with reference to FIG. The slit plate 34 is, for example, a light-transmissive glass plate in which a window on the slit is formed by using chrome or the like, and in the example shown in FIG.
X and SY and fiducial marks MA and MB for each alignment system are formed respectively.

Of these, the light emitting slit SX is formed in a rectangular shape having a width l1 and a length l2, as shown in the projection state on the reticle R in FIG. Here, l1 and l2 are, for example, l1 =
4 μm, 12 = 100 μm.

On the other hand, a cross-shaped reticle mark MR is formed on the reticle R described above, for example, as shown in FIG.
This reticle mark MR is formed of a material having a light shielding property such as chromium.

Next, as shown in FIG. 1, the light LB transmitted through the slit plate 34 on the stage 22 is also transmitted through the projection lens 20, the reticle R, and the condenser lens 18, respectively, and is dichroic mirror.
It is incident on 16 and is reflected here. Then, the reflected light is reflected by the beam splitter 12 via the relay optical system 14, passes through the condenser lens 38, and is a detector for photoelectric conversion.
It is designed to be incident on 40.

The condenser lens 38 is omitted and the beam splitter 12
The detector 40 may be directly arranged after being reflected by.

Here, as will be described in detail later, the detector 40 is arranged at a position that is substantially conjugate with the entrance pupil ep of the projection lens 20, and its light receiving surface has a size that covers the entire pupil ep. It is better to have it.

Further, the numerical aperture formed by the light receiving surface of the detector 40 with respect to the reticle R is the illumination light for exposure as described later.
The characteristics and arrangement of each of the optical elements are set so that LA is equal to the numerical aperture of the reticle R.

Next, the photoelectric output of the detector 40 is amplified by the amplifier 42 and is converted into an analog-digital converter (hereinafter referred to as “ADC”) 4
The data is input to the memory 4, converted into a digital signal here, and input to the memory 46 as data.

The position signal of the interferometer 48 indicating the coordinate position of the stage 22 is input to the ADC 44 and the memory 46, AD conversion is performed at the timing of the position signal, and the output of the ADC 44 is output to the address of the memory 46 represented by the signal. The data is stored synchronously.

Next, the output of the data stored in the memory 46 is performed to the main controller 50, and the main controller 50 drives the stage 22 based on the input data and the output of the focus sensor 30. The motor 52 has a function of outputting a drive signal.

General operation of the embodiment Next, the general operation of the above embodiment will be described. First, when the illumination for exposure is not performed, the light emission slit SX of the slit plate 34 has a rectangular shape as described above, and therefore the rectangular illumination is projected on the reticle R from the lower side of FIG. It will be done through the lens 20.

Then, as the stage 22 moves, the reticle R
The image of the light emission slit SX projected above also moves. In the example of FIG. 4, the light emission slit SX moves in the direction of arrow FB with respect to the reticle mark MR.

By this operation, the image of the light emission slit SX comes to overlap the reticle mark MR. The degree of this overlap is perceived by the detector 40 as a decrease in the amount of light, as shown in FIG. In the example of this figure, it is understood that the light amount is most reduced at the position XB with respect to the positions XA and XC, and the light emission slit SX is well overlapped with the reticle mark MR at this position.

On the other hand, the coordinate position of the stage 22 is detected by the interferometer 48, and position signals are input to the ADC 44 and the memory 46, respectively. In synchronization with this position signal, the photoelectric output of the detector 40 is sampled by the ADC 44, converted into a digital signal, and stored in the corresponding address of the memory 46.

By the way, when the projection lens 20 has no distortion at all, the position where the light emitting slits SX, SY and the reticle mark overlap can be obtained from the design data, and the light amount of the detector 40 stored in this design position and the memory 46 can be reduced. The position or the mark detection point coincides.

However, when the projection lens 20 has distortion, they do not match, and the design position and the mark detection point do not match. The degree of such a difference corresponds to Δ in FIG. 2 (D).

Therefore, the reticle R is covered over the effective range of the projection lens 20
By pre-distributing the pattern of FIG. 4 above, the distortion of the projection lens 20 can be measured. The same applies to the magnification of the projection lens 20.

The movement of the stage for sampling and storing the data as described above is performed under the control of the main controller 50.

Influence of Numerical Aperture (or σ Value) By the way, in this embodiment, the distortion of the projection lens 20 is measured by performing pattern projection in the direction opposite to that at the time of exposure.

Therefore, the numerical aperture or σ value is set so that the distortion of the projection lens 20 that occurs during exposure matches the distortion of the projection lens 20 that occurs during measurement.

Hereinafter, the relationship of the above numerical apertures will be described in detail. FIG. 6A shows the arrangement of the main part and the numerical aperture at the time of exposure. In this figure, the exposure light output from the exposure light source 100 is transmitted through the condenser lens 102, the reticle surface 104 corresponding to the pattern surface of the reticle R, and further through the projection lens 106, and the wafer corresponding to the surface of the wafer W. It is assumed that the light is incident on the surface 108.

Here, the numerical aperture of the exposure light with respect to the reticle surface 104 is NA
_{Let RL be} the numerical apertures of the projection lens 106 with respect to the reticle surface 104 and the wafer surface 108 be NA _R and NA _W , respectively.

Next, FIG. 3B shows the arrangement of the main parts and the numerical aperture at the time of measurement. In this figure, the measurement light source 11
The measurement light output from 0 is transmitted through the condenser lens 112 and the wafer surface 108 to enter the projection lens 106, and the measurement light transmitted therethrough is transmitted through the reticle surface 104 and the condenser lens 102, respectively, and the detector 114. Shall be incident on.

Here, the numerical aperture of illumination of the measurement light on the wafer surface 108 is NA _WL , and the numerical aperture of the detector 114 on the reticle surface 104 is NA _RD . In addition, the slit plate 34 is provided on the wafer surface 108.
Is located and the pattern is projected on the reticle R, as described above.

Further, FIG. 7 shows a coordinate system set in each part of the measurement system as described above. In this example, a one-dimensional coordinate system is used for easy understanding.

As described above, in this embodiment, the change in the light amount of the detector 114 (corresponding to the detector 40 in FIG. 1) is caused by the stage 22,
That is, it is obtained corresponding to the moving position of the slit plate 34 on the wafer surface 108.

Therefore, the amount of light detected by such a detector is determined by the slit plate.
It is determined as a function of the movement amount of 34.

First, the diffraction un
it) (“Fourier imaging theory” (see Teruzo Kose, Kyoritsu Shuppan)), y is the coordinate in the measurement light source 110 as shown in FIG. 7, and A (y) is the intensity distribution of the measurement light source 110. The Fourier component of the amplitude transmittance of the pattern (corresponding to the pattern of the slit plate 34) on the wafer surface 108 is set as exp (iηt) · (η), where η is the Fourier coordinate of the pattern. Yes, t is the amount of movement of the pattern, and (η) is the Fourier component when t = 0.

Light from the position of coordinate y of the measurement light source 110 is emitted from the wafer surface 108.
Then, the light is diffracted in the pattern, and is incident on the projection lens 106 in the x direction. It should be noted that x has a magnitude proportional to sin θ, where θ is the angle between the optical axis and the coordinate indicating the direction.

When light enters the projection lens 106, x = y + η (1) holds. ("Absent condition and OTF calculation") (See Masato Shibuya, Optics 13: No. 1 (1984) PP40-48).

Next, when the magnification of the projection lens 106 is β (= NA _W / NA _R > 0), the direction of the diffracted light of the measurement light emitted from the projection lens 106 is x ′ = x / β (2 ). This light is further diffracted by the Fourier component (ξ) of the pattern on the reticle surface 104, and the detector 11
It is incident on the position of coordinate S on 4.

Here, S = X′−ξ (3) Further, the sensitivity of the detector 114 is S (s) and the pupil function of the projection lens 106 is G (x).

Now, when the incident light intensity distribution I (S, t) on the detector 114 is obtained, the above-described equations (1) to (3) and the measurement light source 110 are calculated.
The intensity distribution A (y) of is an incoherent light source, and I (S, t) = ∫A (y) dy · | ∫ (η) exp (itη) · G (X). (Ξ) d (ξ | ² = ∫A (y) dy · | ∫ (βS + βξ−y) · exp {it (βs + βξ−y)} · G (βs + βξ) · (ξ) dξ | ² Therefore, The total amount of light (t) detected by the detector 114 is (t) = ∫A (y) dy ・ ∫S (s) ds ・ | ∫ (ξ) ・ G (βs + βξ) ・ exp {it (βs + βξ-y) } ((Βs + βξ−y) dξ | ² (4)

Further, the pattern on the wafer surface 108 (slit plate 34
(Corresponding to the pattern) is W (u), it is expressed as (βs + βξ ₁ −y) = C · ∫exp {+ i (βs + βξ ₁ −y) u} W (u) du. Therefore, the equation (4) is expressed as (t) = C · ∫A (y) dy · ∫S (s) ds · ∫dξ ₁ dξ _2. (Ξ ₁ ). ^* (Ξ ₂ ) ・ G (βS + βξ ₁ ) ・ G ^* (βS + βξ ₂ ) ・ exp {itβ (ξ ₁ −ξ ₂ )} ・ ∫du ₁ du ₂・ exp {+ i (βs + βξ ₁ −y) u ₁ } ・exp becomes _{{-i (βs + βξ 2 -y} ) u 2} · W (u 1) · W * (u 2). Here, " ^* " represents a complex conjugate.

By the way, as described above, the numerical aperture of the light LB illuminating the slit plate 34 with respect to the slit plate 34 is configured to be larger than the numerical aperture of the projection lens 20 with respect to the wafer surface.

Thus, if the illumination is performed with a sufficiently large numerical aperture, A (y) = 1 can be set. Therefore, Then, (t) = C · ∫S (s) ds · ∫dξ ₁ dξ ₂ · (ξ ₁ ) · ^* (ξ ₂ ) · G (βs + βξ ₁ ) · G ^* (βs + βξ ₂ ) · ∫du · exp { iβ (t + u) (ξ ₁ −ξ ₂ )} · | W (u) | ² .

Further, when t + u = u ′ is replaced, (t) = C · ∫S (s) ds · ∫dξ ₁ dξ ₂ · (ξ ₁ ) · ^* (ξ ₂ ) · G (βs + βξ ₁ ) · G ^* ( βs + βξ ₂ ) · ∫du ′ · exp {iβu ′ (ξ ₁ −ξ ₂ )} · | W (u′-t) | ² ...
(5) Iw (u ′) = ∫S (s) ds ・ ∫dξ ₁ dξ ₂・ (ξ ₁ ) ・^* (ξ ₂ ) ・ G (βs + βξ ₁ ) ・ G ^* (βs) + Βξ ₂ ) · exp {iβu ′ (ξ ₁ −ξ ₂ )} (6), Iw (u ′) is the wafer surface 108 of the object on the reticle surface 104 illuminated by the equivalent light source S (s). It is the intensity at the upper point u '.

From equations (5)-(6), (t) = C · ∫du ′ · Iw (u ′) · | W (u′-t) | ² (7) is obtained. The expression (7) is a convolution of the pattern on the wafer surface 108 when the pattern on the wafer surface 108 moves t and the intensity distribution of the image on the reticle surface 104.

By the way, since | W (u′−t) | ² is known, Iw (u ′) can be obtained from the above-mentioned equation (7). If the shape of the detector 114 is the same as the shape of the equivalent light source at the time of exposure, that is, the numerical aperture NA _RD of the detector 114 with respect to the reticle surface 104 shown in FIG. When the numerical aperture NA _{RL of the} exposure light source (100) with respect to the reticle surface 104 is equal, the projection lens 10 at the time of exposure
At 6, the imaging characteristics will be measured by the detector 114.

It should be noted that, in other words, the above conditions correspond to making the σ values on the reticle surface 104 the same during measurement and during exposure. Further, in the above description, the numerical aperture NA of the measuring illumination is
_It is assumed that _WL (see FIG. 6 (B)) is sufficiently large, but if it is larger than the numerical aperture (NA _W (see FIG. 6)) of the projection lens 106, the zero-order diffracted light of the measurement light passes through the projection lens 106. As a result, almost satisfactory results are obtained.

This is because if the so-called σ value of illumination is σ> 1, the image forming performances are almost the same (see, for example, “Fourier image formation theory” Teruji Kose, Kyoritsu Shuppan).

Furthermore, when a self-luminous object is used as the pattern on the wafer surface 108 corresponding to the slit plate 34, it corresponds to the above-mentioned numerical aperture NA _WL being sufficiently large. Good.

As described above, the numerical aperture NA _RD that the detector 114 makes with respect to the reticle surface 104 via the condenser lens 102.
Is equal to the numerical aperture NA _RL that the exposure light source 100 makes with respect to the reticle surface 104 through the condenser lens 102 during exposure, and the numerical aperture NA _WL that the measurement light source 110 makes with the wafer surface 108 through the condenser lens 112 is If it is larger than the numerical aperture NA _W of the projection lens 106, the distortion generated in the projection lens 106 at the time of exposure and the measured distortion satisfactorily match, and the measurement can be performed accurately.

By the way, the detector 114 (corresponding to the detector 40 in the example of FIG. 1) is optically arranged substantially conjugate with the pupil of the projection lens 106, but the detector 114 and the reticle surface 104 are optically arranged. The arrangement may be conjugate.

However, if there is uneven sensitivity of the detector 114,
Since there is a measurement error in the arrangement that is conjugate with the reticle surface 104,
It is preferable to use a conjugate arrangement with the pupil of the projection lens 106.

Therefore, in the embodiment shown in FIG. 1, the detector 40 is arranged so as to be conjugate with the entrance pupil ep of the projection lens 20 so as to be substantially conjugate.

Further, in FIGS. 6 to 7, the measurement light source 110 and the pupil of the projection lens 106 are arranged so as to be substantially conjugated, but the measurement light source 110 and the wafer surface 108 may be conjugated,
The condenser lens 112 may be omitted.

However, in such a case, the intensity ram of the measurement light source 110 causes uneven illumination intensity on the wafer surface 108. Therefore, it is better to make the measurement light source 110 substantially conjugate with the pupil of the projection lens 106. This is because Koehler illumination is less affected by light source unevenness as compared with critical illumination, but other characteristics are the same (for example, “Principle of optics” (Born and
Wolf, Kusagawa, Yokota translation, Tokai University Press, P.781-786).

In the embodiment shown in FIG. 1, the end face of the fiber 34D is arranged so as to be substantially conjugate with the exit pupil (not shown) of the projection lens 20.

When the σ value during exposure illumination is changed, the σ
You can change the value. For example, in the embodiment shown in FIG. 1, the fly-eye exit surface 10 is substantially conjugate with the pupil of the projection lens 20, and the pupil is also conjugated with the detector 40. When a diaphragm is placed and its size is changed to change the σ value, the diaphragm is placed on the detection surface of the detector 40 and the size is also changed, so that the measured σ value is changed to the exposure σ value.
Can match the value. In principle, the sizes of the two diaphragms may be changed in a proportional relationship, but it is necessary to correct depending on the aberration of the condenser lens 38, but the amount can be given as an offset amount at the time of design.

Overall Operation of Embodiment Next, the overall operation of the above embodiment will be described based on the above points.

First, without exposure, the slit plate 34 is illuminated with the detection light LB for distortion measurement from the inside of the stage.

As a result, the reticle R is illuminated with a rectangular shape from below in FIG. 1 via the projection lens 20.

Next, when the stage 22 is moved by the drive command to the motor 52 of the main controller 50, the light emission slit image projected on the reticle R also moves and overlaps with the reticle mark MR.

The degree of this overlap is detected as a decrease in the amount of incident light on the detector 40, as shown in FIG.

This detection output is AD converted and stored in the memory 46, as described above. The distortion of the projection lens 20 is measured using such data and the data at the time of design.

By the way, as described above, the numerical aperture of the detector 40 with respect to the reticle R at the time of measurement and the numerical aperture of the exposure light with respect to the reticle R at the time of exposure are set to be equal, and the slit of the measuring illumination light is set. The numerical aperture for the plate 34 is set larger than the numerical aperture for the wafer surface of the projection lens 20.

Therefore, the measured image formation characteristic of the projection lens 20 and the image formation characteristic of the projection lens 20 at the time of exposure satisfactorily coincide with each other, and accurate measurement is performed.

Effect of Embodiment As described above, according to this embodiment, there is an effect that the distortion and the magnification can be accurately measured in a short time.

In addition, since the measured data is stored inside the device, when the position of the alignment mark on the wafer and the position of the alignment mark on the reticle aligned with this mark are different, the alignment error due to distortion may occur. There is an effect that the minute can be easily corrected.

Other Embodiments The present invention is not limited to the above embodiments, and a semiconductor laser, for example, may be provided in the stage (for example, Z stage 22A). In this way, the fiber 34D and the like can be omitted.

Further, the position of the slit plate 34 and the position of the surface of the wafer W do not necessarily have to coincide with each other, and it is sufficient that the position is on the image forming plane of the projection lens 20 at the time of measurement.

Further, since the detector 40 can also receive the reflected light of an object (for example, a wafer) located on the image plane of the projection lens 20, it is possible to measure the reflectance of the object.

Although the distortion and the magnification are often measured in both directions of the XY plane, they may be measured only in the one-dimensional direction, if necessary.

[Effects of the Invention] As described above, according to the present invention, by using the light emitting means provided on the stage and the mark means formed on the mask, the numerical aperture of the exposure light with respect to the mask and the detector. While matching the numerical aperture for the mask,
Since the numerical aperture for the light emitting means of the measurement illumination is set to be larger than the numerical aperture for the projection target of the projection optical system,
There is an effect that distortion and magnification can be accurately measured in a short time.

[Brief description of drawings]

FIG. 1 is a structural view showing an embodiment of the present invention, FIG. 2 is a partial structural view showing a detailed structural example of a stage portion in the above embodiment, and FIG. 3 is a plan view showing an example of a slit plate. FIG. 4 is an explanatory view of the overlap between the reticle mark and the light emitting slit, FIG. 5 is a diagram showing a waveform example of the photoelectric signal of the detector, and FIG. 6 shows the arrangement and the numerical aperture of the main part at the time of exposure and measurement. Explanatory diagram, FIG. 7 is an explanatory diagram showing coordinates set in a main part at the time of measurement, and FIG. 8 is an explanatory diagram of a conventional measuring method. [Explanation of symbols of main parts] 20 ... Projection lens, 22 ... Stage, 34 ... Slit plate, 40 ... Detector, 46 ... Memory, 48 ... Interferometer,
50 ... Main control unit, 52 ... Motor, MR ... Reticle mark, R ... Reticle, W ... Wafer, SX, SY ... Light emission slit, NA _R , NA _RD , NA _RL , NA _W , NA _WL …… Numerical aperture.

Claims (1)

(57) [Claims] 1. A projection optical apparatus for illuminating a mask on which a desired pattern is formed with illumination light from an illumination optical system to form an image of the pattern on a predetermined image plane via the projection optical system. A stage capable of moving along the image plane of the projection optical system, a stage position detecting means for outputting a position signal according to the position of the stage, and a light emitting means provided on the stage and having a light emitting surface of a predetermined shape. A photoelectric detection means for receiving light from the light emitting surface of the light emitting means through the projection optical system and a mask, and a projected image of the light emitting surface with respect to a mark pattern formed at a predetermined position of the mask. The stage is controlled so as to move, and the projection of the light emitting surface is performed based on a photoelectric signal output from the photoelectric detection unit during the movement and a position signal output from the stage position detection unit. Wherein a position detecting means for detecting a positional relationship between the mask pattern, the numerical aperture with respect to the mask of the light receiving surface of the photoelectric detection means and,
The projection optical device is characterized in that the numerical aperture of the illumination light is set equal to the numerical aperture of the mask, and the σ value of the illumination by the light emitting means is set to 1 or more.