JPH0645217A - Projection exposure device - Google Patents

Abstract

(57) [Abstract] [Purpose] Even if the illumination conditions for the reticle are changed, the wafer is always exposed under highly accurate imaging characteristics. [Arrangement] When the illumination conditions (σ value, annular illumination, multiple tilt illumination, etc.) for the reticle R are changed according to the type of the reticle R and the fineness of the pattern, the image forming characteristics of the projection optical system are changed. Until the amount of change falls below a predetermined allowable value,
The exposure operation for the wafer W is stopped.

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 for manufacturing semiconductor integrated circuits and liquid crystal devices, which requires high-precision image forming performance, and more particularly to maintaining the image forming performance of a projection optical system.

[0002]

2. Description of the Related Art In a photolithography process for forming a circuit pattern of a semiconductor element or the like, a method of transferring a pattern formed on a reticle (mask) onto a substrate (semiconductor wafer, glass plate, etc.) is usually adopted. It A photosensitive photoresist is coated on the substrate, and the circuit pattern is transferred to the photoresist according to the irradiation light image, that is, the pattern shape of the transparent portion of the reticle pattern. In a projection exposure apparatus (for example, a stepper or the like), an image of a reticle pattern is image-projected on a wafer via a projection optical system.

In this type of apparatus, in the plane of the illumination optical system (hereinafter referred to as the pupil plane of the illumination optical system), which is the Fourier transform plane of the plane on which the pattern on the reticle exists, or in the plane in the vicinity thereof, The illumination light flux is limited to a substantially circular shape (or a rectangular shape) about the optical axis of the illumination optical system to illuminate the reticle. Therefore, the illumination light flux is incident on the reticle at an angle that is almost vertical. Further, on a reticle (a glass substrate such as quartz) used in this apparatus, a transmissive portion (reticle bare surface portion) having a transmittance of almost 100% for an illumination light flux and a light shielding portion having a transmittance of almost 0%. A circuit pattern composed of (chrome etc.) was drawn.

Now, the illumination light beam irradiated on the reticle as described above is diffracted by the reticle pattern, and 0th order diffracted light and ± 1st order diffracted light are generated from the pattern. These diffracted lights are condensed by the projection optical system and an interference fringe, that is, an image of a reticle pattern is formed on the wafer. At this time, the angle θ (reticle side) formed by the 0th-order diffracted light and the ± 1st-order diffracted light is sin θ = λ / P, where λ (μm) is the wavelength of the exposure light and NA is the reticle-side numerical aperture of the projection optical system. Determined by

By the way, if the pattern pitch becomes finer,
When sin θ becomes larger and sin θ becomes larger than the reticle-side numerical aperture (NA) of the projection optical system, the ± 1st-order diffracted light becomes the Fourier transform surface of the reticle pattern. It is limited by the effective diameter of the surface) and cannot pass through the projection optical system. In other words, only the 0th-order diffracted light reaches the wafer, and no interference fringe (pattern image) is generated. Therefore, in the conventional exposure method as described above, when a reticle consisting of only the above-mentioned transmissive portion and light-shielding portion (hereinafter referred to as a normal reticle) is used, the fineness of the reticle pattern that can be resolved on the wafer (minimum pattern pitch ) P is P ≈ P from sin θ = NA
It is given by the relational expression λ / NA.

Therefore, the minimum pattern size is the pitch P.
Therefore, the minimum pattern size is 0.5 × λ /
Although it is about NA, in the actual photolithography process, the influence of the wafer curvature, the step difference of the wafer due to the process,
Alternatively, some depth of focus is required due to the thickness of the photoresist itself. Therefore, the practical minimum resolution pattern size is expressed as k × λ / NA. Here, k is called a process coefficient and is usually about 0.6 to 0.8.

From the above, in the conventional exposure method
Shorter for exposing and transferring finer patterns
Wavelength exposure light source, or higher numerical aperture
It was necessary to use a fine projection optical system. But Naga
Have a shorter wavelength of the exposure light source (for example, 200 nm or less).
Below) is suitable for use as a transmissive optical member.
Absence of optical material, stable with high light output
It is difficult at this point due to the lack of a light source.
Also, the numerical aperture of the projection optical system is already at the theoretical limit.
It is almost hopeless that a larger aperture will be created soon. Tentative
± λ / NA even if a larger aperture than the shape is possible ²
The depth of focus, which is determined by, decreases sharply as the numerical aperture increases.
Therefore, the depth of focus required for actual use will be smaller,
There is a problem that it cannot be a general purpose exposure apparatus.

Therefore, a phase shifter (dielectric thin film) for shifting the phase of the transmitted light from a specific portion of the transmitted portion of the circuit pattern of the reticle by π (rad) with respect to the transmitted light from other transparent portions. It has also been proposed to use a phase shift reticle with The phase shift reticle is disclosed in, for example, Japanese Patent Publication No. 62-50811, and the use of this phase shift reticle makes it possible to transfer a finer pattern than in the case of using a normal reticle. That is, there is an effect of improving the resolution. When using this phase shift reticle,
The numerical aperture (coherence factor σ) of the illumination optical system needs to be optimized. Various methods have been proposed so far for the phase shift reticle, but typical ones are the spatial frequency modulation type, the shifter shading type, and the edge emphasis type.

[0009] Recently, attempts have been made to enable transfer of a fine pattern by optimizing illumination conditions or devising an exposure method. For example, an aperture of an illumination optical system that is optimal for a pattern having a specific line width. There has been proposed a method of improving resolution and depth of focus by selecting a combination of a numerical value (σ value) and a numerical aperture (NA) of a projection optical system for each pattern line width. Furthermore, the light quantity distribution of the illumination light flux on the pupil plane of the illumination optical system or in the vicinity of the pupil plane is regulated in a ring-shaped manner, and the ring-shaped illumination method that illuminates the reticle pattern with the illumination light flux, or the periodicity of the reticle pattern is supported. There is also proposed an inclined illumination method or the like in which an illumination light beam is inclined from a specific direction and emitted at a predetermined angle. However, in any of the above methods, it is not effective for all reticle patterns, that is, the line width and shape thereof, and it is necessary to select an optimum illumination method or condition for each reticle or its pattern. The projection exposure apparatus requires a structure that allows the illumination conditions (σ value, etc.) in the illumination optical system to be variable.

By the way, in recent years, projection exposure apparatuses are increasingly required to maintain the imaging characteristics (projection magnification, focus position, etc.) of the projection optical system at a constant value with high accuracy. Therefore, various methods of correcting the imaging characteristics have been proposed and put into practical use. Among them, a method for correcting the fluctuation of the image forming characteristic due to the absorption of the exposure light of the projection optical system is disclosed in, for example, JP-A-60-78454. In the disclosed method, the amount of energy (heat amount) accumulated in the projection optical system with the incidence of exposure light (i-line, KrF excimer laser, etc.) on the projection optical system is sequentially calculated, and the accumulated energy amount is used. It has been proposed to obtain the amount of change in the image forming characteristic and finely adjust the image forming characteristic by a predetermined correction mechanism. As the correction mechanism, for example, two of the plurality of lens elements that form the projection optical system are used.
There is a method of sealing the space sandwiched between two lens elements and adjusting the pressure in this sealed space.

[0011]

However, in the prior art as described above, when the illumination condition of the illumination optical system is changed as described above or when the phase shift reticle is used, the lens near the pupil plane of the projection optical system is used. In the element, the distribution of the amount of transmitted light can change. Since the illumination light originally concentrates and passes in the vicinity of the pupil plane, a change in the light amount distribution here has a great influence on the variation of the imaging characteristics due to the illumination light absorption of the projection optical system. Therefore, as disclosed in, for example, Japanese Unexamined Patent Publication No. 62-229838, the calculation parameter used for calculating the amount of change in the image forming characteristic of the projection optical system due to the absorption of illumination light is corrected for each illumination condition. It is conceivable that the correction parameter is used to accurately obtain the change in the imaging characteristic due to the change in the illumination condition and the correction is performed.

When correcting the image forming characteristics as described above, there is no problem from a long-term standpoint. However, the phenomenon of heat accumulation in the projection optical system has a past history. Therefore, when the illumination condition is changed in accordance with the reticle or its pattern, if the amount of change in the imaging characteristics is immediately calculated and corrected with the calculation parameter corrected under the new illumination condition, the previous illumination is changed. Since the history corresponding to the condition remains in the projection optical system, inconvenience arises in correcting the imaging characteristics while the history remains. That is, immediately after changing the illumination condition, the heat distribution state under the previous illumination condition and the heat distribution state under the new illumination condition overlap in the lens element near the pupil plane of the projection optical system. It is in a state where there is neither heat distribution, and no matter which illumination condition is used to calculate the change amount of the imaging characteristics with the calculation parameters, the calculation result is the same as the actual change amount of the imaging characteristics. It does not match. The image formation characteristic in such a transient state (that is, the heat distribution state of the projection optical system) cannot be simply expressed as the sum of the two, and the change amount of the image formation characteristic in this transitional state can be calculated accurately. , Very difficult to correct. Therefore, even if pattern exposure is performed on the wafer under the above-mentioned transitional state, it is not possible to obtain a circuit pattern that satisfies the desired characteristics.

The present invention has been made in consideration of the above points, and the illumination conditions (numerical aperture (σ value of the illumination optical system) of the reticle and the reticle pattern (type, periodicity, fineness, etc.) are taken into consideration. ), The numerical aperture of the projection optical system, annular illumination, tilted illumination, etc.) can be changed to obtain a projection exposure apparatus that can always perform pattern exposure on a photosensitive substrate under highly accurate imaging characteristics. It is an object.

[0014]

In order to solve such a problem, in the present invention, the illumination light IL from the light source 1 is shaped into a substantially uniform intensity distribution, and the uniform illumination light I
An illumination optical system (1 to 12) for irradiating a reticle R having a fine pattern with L, a projection optical system PL for projecting an image of the pattern of the reticle R on a wafer W, and a projection optical system PL.
Means for controlling the image forming characteristics of the (driving elements 25, 27, 2
9. In the projection exposure apparatus including the drive element control unit 53), the type of reticle (normal reticle, phase shift reticle) or the fineness of the pattern (periodicity, pitch,
Illumination condition changing means (variable aperture diaphragms 8, 32, variable blind 10, or turret plate 7 and drive system 5) for changing the illumination condition for the reticle R in accordance with the direction, etc.
4) and; exposure control means (main control) that stops the exposure operation on the wafer W for a predetermined time (that is, the time until the amount of change ΔM of the imaging characteristics reaches a predetermined level M _S1 ) when the illumination condition is changed. And the device 50).

[0015]

According to the present invention, when the illumination condition is changed in accordance with the mask or the pattern thereof, the change amount of the image formation characteristic due to the illumination light absorption of the projection optical system under the illumination condition before the change is predetermined. After the exposure operation on the photosensitive substrate is stopped until it becomes within the allowable value of, the exposure operation is performed under the new illumination condition while controlling the image forming characteristic of the projection optical system. Therefore, since the exposure operation is not performed under the transitional state when the illumination condition is changed, it is possible to strictly control the image forming characteristic of the projection optical system for each illumination condition.

[0016]

1 is a plan view showing a schematic structure of a projection exposure apparatus according to an embodiment of the present invention. In FIG. 1, the ultra-high pressure mercury lamp 1 generates illumination light (i-line or the like) IL in a wavelength range that exposes the resist layer. As the exposure illumination light source 1, a bright line such as a mercury lamp, a laser light source such as a KrF or ArF excimer laser, or a harmonic such as a metal vapor laser or a YAG laser may be used. The illumination light IL is reflected by the elliptical mirror 2 and condensed at the second focal point f ₀ thereof, and then is passed through a cold mirror 5 and a condensing optical system 6 including a collimator lens and the like to an optical integrator (fly-eye lens group) 7A. Incident on. A variable aperture stop 8 for varying the numerical aperture NA _IL of the illumination optical system is arranged near the exit surface (reticle side focal plane) of the fly-eye lens group 7A. Here, the fly-eye lens group 7A is
The focal plane on the reticle side is arranged in the in-plane direction perpendicular to the optical axis AX so that the focal plane on the reticle substantially coincides with the Fourier transform plane (pupil conjugate plane) of the reticle pattern. In the vicinity of the second focal point f ₀ , the motor 4 closes the optical path of the illumination light IL,
A shutter (for example, a four-blade rotary shutter) 3 for opening is arranged.

In the present embodiment, a plurality of fly-eye lens groups (FIG. 2) including a fly-eye lens group 7A are provided on the holding member (rotary turret plate) 7, and a drive system 54 allows an arbitrary fly-eye lens group. The lens group is configured to be replaceably arranged in the illumination optical path. Therefore,
By configuring the plurality of fly-eye lens groups to be replaceable together with the variable aperture stop 8, it is possible to change the illumination condition according to the type of reticle, the periodicity of the pattern, etc. (details will be described later).

The illumination light IL emitted from the fly-eye lens group 7A is relay lenses 9 and 11, a variable blind 10,
And the mirror 1 passing through the main condenser lens 12.
3, the pattern area PA of the reticle R is illuminated with a substantially uniform illuminance after being reflected almost vertically downward. Since the surface of the variable blind 10 has a conjugate relationship with the reticle R, the variable blind 10 can be moved by a motor (not shown).
The illumination field of the reticle R can be arbitrarily set by opening and closing the plurality of movable blades constituting the above to change the size and shape of the opening. The reticle R is held by the reticle holder 14, and the holder 14 is placed on the reticle stage RS which can be two-dimensionally moved in the horizontal plane by a plurality of expandable and contractible drive elements 29 (only two are shown in FIG. 1). . Therefore, by controlling each expansion / contraction amount of the drive element 29 by the drive element control unit 53, the reticle R can be translated in the optical axis direction and can be tilted in an arbitrary direction with respect to a plane perpendicular to the optical axis. Has become. As will be described in detail later, the above-described configuration can correct the image forming characteristics of the projection optical system, in particular, the pincushion type and barrel type distortions. The reticle R is positioned so that the center point of the pattern area PA coincides with the optical axis AX.

The illumination light IL that has passed through the pattern area PA enters the projection optical system PL that is telecentric on both sides, and the projection optical system PL forms a projected image of the circuit pattern of the reticle R on which a resist layer is formed. Projection (imaging) is performed by superimposing it on one shot area on the wafer W held so that its surface is substantially coincident with the best imaging plane. still,
In this embodiment, it is possible to independently drive some of the lens elements (20, 21, and 22 in the figure) that form the projection optical system PL, and the imaging characteristics of the projection optical system PL, For example, projection magnification, distortion, field curvature, astigmatism, etc. can be corrected (details will be described later).
Further, a variable aperture stop 32 is provided in the pupil plane Ep of the projection optical system PL or in a plane in the vicinity thereof, and the numerical aperture NA of the projection optical system PL can be changed by this.

The wafer W is a wafer holder (θ table) 1
6 is vacuum-sucked and held on the wafer stage WS via the holder 16. The wafer stage WS is
The motor 17 can be tilted in any direction with respect to the best image plane of the projection optical system PL, can be finely moved in the optical axis direction (Z direction), and can be two-dimensionally moved by a step-and-repeat method. When the transfer exposure of the reticle R onto one shot area on the wafer W is completed, the stepping is performed to the next shot position. still,
The structure of the wafer stage WS is disclosed in, for example, Japanese Patent Laid-Open No. 62-274201. A movable mirror 19 that reflects the laser beam from the interferometer 18 is fixed to the end of the wafer stage WS, and the two-dimensional position of the wafer stage WS is determined by the interferometer 18 with a resolution of, for example, about 0.01 μm. Always detected.

Further, both the reference member 15 and the irradiation amount monitor (photoelectric sensor) 33 are provided on the wafer stage WS so as to substantially coincide with the surface position of the wafer W.
On the surface of the reference member 15, a grid pattern used to detect a focal position at an arbitrary position within the image field of the projection optical system PL is formed. The structure of the lattice pattern, the detection operation, etc. will be described in detail later. The photoelectric sensor 33 is composed of, for example, a photodetector having a light-receiving surface having substantially the same area as the image field of the projection optical system PL or the projection area of the reticle pattern, and outputs optical information regarding this irradiation amount to the main controller 50. To do. This light information serves as basic data for obtaining the amount of change in the image forming characteristics of the projection optical system PL.

Further, in FIG. 1, an image forming light beam for forming an image of a pinhole or a slit toward the best image forming surface of the projection optical system PL is supplied obliquely with respect to the optical axis AX. Irradiation optical system 30 and wafer W of the image forming light beam
An oblique-incidence type focus detection system including a light-receiving optical system 31 that receives a light beam reflected by the surface of the light via the beam is provided. The configuration of this focus detection system is disclosed in, for example, Japanese Patent Laid-Open No. 60-168112, and the position of the wafer surface in the vertical direction (Z direction) with respect to the image plane is detected to detect the wafer W and the projection optical system. The in-focus state with PL is detected. In this embodiment, the angle of the parallel flat glass (plane parallel) (not shown) provided inside the light receiving optical system 31 is adjusted in advance so that the image plane becomes the zero reference, and the focus detection system is calibrated. Shall be performed. Further, for example, a horizontal position detecting system as disclosed in Japanese Patent Laid-Open No. 58-113706 is used, or the focal position can be detected at arbitrary plural positions in the image field of the projection optical system PL. It is assumed that the focus detection system is configured (for example, a plurality of slit images are formed in the image field) so that the inclination of a predetermined region on the wafer W with respect to the image plane can be detected.

By the way, in FIG. 1, a main controller 50 for integrally controlling the entire apparatus and a bar code BC representing a name formed on the side of the reticle pattern while the reticle R is being conveyed right above the projection optical system PL. To drive a bar code reader 52 for reading, a keyboard 51 for inputting commands and data from an operator, and a holding member (rotary turret plate) 7 to which a plurality of fly-eye lens groups including a fly-eye lens group 7A are fixed. Drive system (motor, gatoren, etc.) 54 of FIG. In the main controller 50, the names of a plurality of reticles to be handled by this projection exposure apparatus (eg stepper) and the stepper operation parameters corresponding to each name are registered in advance. When the bar code reader 52 reads the reticle bar code BC, the main control unit 50 pre-registers the illumination condition (type of reticle, periodicity of the reticle pattern, etc.) as one of the operation parameters corresponding to the name. (Corresponding to (1)), one fly-eye lens group most suitable for (1) is selected from the holding members 7, and a predetermined drive command is output to the drive system 54. Further, as operating parameters corresponding to the above names, optimum setting conditions of the variable aperture diaphragms 8 and 32 and the variable blind 10 under the fly-eye lens group selected previously, and the image forming characteristics of the projection optical system PL. A calculation parameter (details will be described later) used for correcting by the correction mechanism described later is also registered, and these condition settings are also performed at the same time when the fly-eye lens group is set. As a result, the optimum illumination condition is set for the reticle R placed on the reticle stage RS. The above operation can also be executed by the operator directly inputting commands and data to the main controller 50 from the keyboard 51.

Next, a mechanism for correcting the image forming characteristic of the projection optical system PL will be described. As shown in FIG. 1, in the present embodiment, the driving element controller 53 drives each of the reticle R and the lens elements 20, (21, 22) independently to correct the imaging characteristics of the projection optical system PL. It is possible. As the image forming characteristics of the projection optical system PL,
There are focal position, projection magnification, distortion, field curvature, spherical aberration, astigmatism, etc., and these values can be individually corrected, but in this embodiment, in order to simplify the description, in particular, both sides A case of correcting the focal position, the projection magnification, the distortion, and the curvature of field in the telecentric projection optical system will be described. In this embodiment, the barrel-type or pincushion-type distortion is corrected by moving the reticle R.

The first group of lens elements 20 closest to the reticle R is fixed to the supporting member 24, and the second group of lens elements (21, 22) is fixed to the supporting member 26. Lens elements below the lens element 23 are fixed to the lens barrel portion 28 of the projection optical system PL. In the present embodiment, the optical axis AX of the projection optical system PL
Is the optical axis of the lens element fixed to the lens barrel portion 28.

A plurality of support members 24 (for example, 3
Then, two driving elements 25 (shown in the drawing) are connected to the supporting member 26, and the supporting member 26 is connected to the lens barrel portion 28 by a plurality of expandable and contractable driving elements 27.
The drive elements 25, 27, 29 are, for example, electrostrictive elements,
A magnetostrictive element is used, and the displacement amount of the drive element according to the voltage or magnetic field applied to the drive element is obtained in advance. Although not shown here, in consideration of the hysteresis of the drive element, a position detector such as a capacitive displacement sensor or a differential transformer is provided in the drive element, and the position of the drive element corresponding to the voltage or the magnetic field applied to the drive element is provided. Is monitored to enable high-precision driving.

Here, the lens elements 20, (21,
When each of 22) is translated in the optical axis direction, each of the projection magnification M, the field curvature C, and the focus position F changes at a change rate corresponding to the amount of movement. Assuming that the driving amount of the lens element 20 is x ₁ and the driving amount of the lens element (21, 22) is x ₂ , the projection magnification M, the field curvature C, and the change amounts ΔM, ΔC, and ΔF of the focus position F are , Is expressed by the following equation.

ΔM = C _M1 × x ₁ + C _M2 × x ₂ (1) ΔC = C _C1 × x ₁ + C _C2 × x ₂ (2) ΔF = C _F1 × x ₁ + C _F2 × x ₂ (3) C _M1 , C _M2 , C _C1 , C _C2 , C _F1 , and C _F2 are constants that represent the rate of change of each change amount with respect to the drive amount of the lens element.

By the way, as described above, the focus detection system 30,
Reference numeral 31 is for detecting the amount of deviation of the wafer surface from the optimum focus position with the optimum focus position of the projection optical system as a zero point reference. Therefore, when an appropriate electrical or optical offset amount x ₃ is applied to the focus detection systems 30 and 31, the wafer surface is positioned by using the focus detection systems 30 and 31, and the lens element 20 is thus positioned. ,
It is possible to correct the focus position shift due to the driving of (21, 22). At this time, the above expression 3 is expressed as the following expression.

ΔF = C _F1 × x ₁ + C _F2 × x ₂ + x ₃ (4) Similarly, when the reticle R is translated in the optical axis direction, the distortion D and the focus position are changed at a rate corresponding to the amount of movement. Each of F changes. Drive amount of reticle R is x
_{If 4} , the distortion D and the change amounts ΔD and ΔF of the focus position F are expressed by the following equations.

ΔD = C _D4 × x ₄ (5) ΔF = C _F1 × x ₁ + C _F2 × x ₂ + x ₃ + C _F4 × x ₄ (6) Note that C _D4 and C _F4 are the reticle R of each variation. It is a constant indicating the rate of change with respect to the drive amount. From the above, Equation 1,
By setting the drive amounts x _{1 to} x ₄ in 2, 5, and 6, the change amounts ΔM, ΔC, ΔD, and ΔF can be arbitrarily corrected. Although the case where four types of image forming characteristics are simultaneously corrected is described here, if the change amount of the image forming characteristics of the projection optical system due to absorption of illumination light is negligible, It is not necessary to perform the above correction, and when the imaging characteristics other than the four types described in the present embodiment change significantly, it is necessary to perform the correction for the imaging characteristics. Further, in the present embodiment, an example in which correction is performed by moving the reticle R and the lens element as the image formation characteristic correction mechanism has been shown, but a suitable correction mechanism in this embodiment may be any other method, for example, two A method of sealing the space sandwiched between the lens elements and adjusting the pressure of this sealed space may be adopted.

Here, in the present embodiment, the drive element control section 53
The reticle R and the lens element 20,
Although (21, 22) is movable, in particular, the lens elements 20, (21, 22) have an influence on the characteristics such as projection magnification, distortion, field curvature, and astigmatism as compared with other lens elements. Greatly easy to control. In this embodiment, the movable lens element is composed of two groups, but it may be composed of three or more groups. In this case, the range of movement of the lens element can be increased while suppressing the fluctuation of other aberrations, and various lens elements can be used. Shape distortion (trapezoid,
It is possible to deal with distortion such as rhombus), field curvature, and astigmatism. By adopting the correction mechanism having the above-mentioned configuration, it is possible to sufficiently cope with a change in the image forming characteristic of the projection optical system PL due to absorption of exposure light.

With the above configuration, the drive element control section 53
Can independently move the lens elements 20, (21, 22) of the second group and the peripheral points 3 to 4 of the reticle R in the optical axis direction by an amount corresponding to the drive command given from the main controller 50. As a result, the second group of lens elements 20,
Each of (21, 22) and the reticle R can be moved in parallel in the optical axis direction and can be tilted in any direction with respect to the plane perpendicular to the optical axis AX.

Next, the measuring mechanism of the image forming characteristic of the projection optical system PL will be described with reference to FIG. As shown in FIG. 7B, a grid pattern 15m including two sets of slit patterns Gx and Gy is formed on the surface of the reference member 15, and the grid pattern 15m includes the mirror 60 and the lens 6.
1. The illumination light transmitted by the optical fiber 62, the lens 63, and the mirror 64 is illuminated from below.
It is desirable that this illumination light is light in the substantially same wavelength range as the exposure illumination light IL, and for example, a part of the illumination light IL from the light source 1 may be branched and guided by a beam splitter. The illumination light transmitted through the reference member 15 reaches the reticle R via the projection optical system PL, is reflected by the back surface (pattern surface) thereof, and returns to the reference member 15 again via the projection optical system PL. Further, the reflected light transmitted through the grating pattern 15m again passes through the optical fiber 62 and the like, and then enters the photodetector 65 composed of Photomal or SPD, where it is photoelectrically converted and a photoelectric signal corresponding to the light intensity is converted into a photoelectric signal. Output. The main controller 50 inputs, for example, a detection signal from the focus detection system (light receiving optical system 31) together with the photoelectric signal from the photodetector 65. Then, the wafer stage WS
Is moved in the Z direction, and the level change of the photoelectric signal of the photodetector 65 at that time is obtained. The result is shown in FIG. In the figure, the position f at which the signal level has a peak is the optimum focus position (best focus position) at any one point in the image field. For example, the measurement is performed while moving the reference member 15 in the image field. By repeatedly performing, the focus position, field curvature, astigmatism, etc. can be obtained as the imaging characteristics of the projection optical system PL. Since the projection magnification and the distortion cannot be measured with the above configuration, for example, the marks formed at a plurality of positions on the reticle R are photoelectrically detected via the reference marks provided on the wafer stage. It is desirable to keep.

Although not shown in FIG. 7A,
In this embodiment, the exit surface of the optical fiber 62 (on the side of the reference member)
It is assumed that a plurality of spatial filters can be exchangeably arranged in the vicinity. Accordingly, the illumination condition for the reference member 15 (for example, the numerical aperture (σ of the optical fiber 62 (σ
Value), orbicular zone illumination, plural-tilt illumination described later, etc.) can be arbitrarily changed. At this time, it is desirable to provide slit patterns with a plurality of pitches on the surface of the reference member 15 in correspondence with the reticle pattern.

Next, the illumination method adopted in the projection exposure apparatus according to the present embodiment will be described with reference to FIGS. As an example of the illumination method adopted in this embodiment, the illumination light flux passing through the pupil plane of the illumination optical system, its conjugate plane, or the plane in the vicinity thereof is decentered by a predetermined amount from the optical axis AX of the illumination optical system. By defining at least two local areas having the center at the position, the illumination light flux irradiated to the reticle R is tilted in a predetermined direction by an angle according to the fineness of the reticle pattern (hereinafter, simply referred to as a plurality of Called the inclined illumination method).

FIG. 2 is a diagram showing how diffracted light is generated from a circuit pattern and an image is formed when a reticle is illuminated by using the multiple tilt illumination method. On the lower surface of the reticle R (on the side of the projection optical system PL), a one-dimensional line-and-space pattern including a transmissive portion Ra and a light shielding portion Rb is drawn as a circuit pattern. In the projection exposure apparatus (FIG. 1) used in this embodiment, the local region through which the illumination light flux passes has a center at a position decentered from the optical axis in the pupil plane of the illumination optical system, as will be described later. There is. Therefore, the illumination light flux L0 that illuminates the reticle R is
The light is incident on the reticle R at a predetermined incident angle ψ from a direction (X direction) substantially perpendicular to the direction in which the circuit pattern of (1) is drawn (period direction). The incident angle ψ and the incident direction are uniquely determined by the position of the diffracted light generated by the pattern on the reticle in the pupil plane of the projection optical system.

From the pattern on the reticle, 0th-order diffracted light _Do , + 1st-order diffracted light _Dp , and -1st-order diffracted light _Dm are generated in the direction of the diffraction angle according to the fineness (width, pitch) of the pattern. In FIG. 2, what is transmitted through the projection optical system PL and reaches the wafer W is the 0th-order diffracted light D _o and +1 among the above three light fluxes.
It is shown as two light beams of the next diffracted light D _p , and these two light beams form an interference fringe, that is, an image of a circuit pattern on the wafer W. The positions of the 0th-order diffracted light D _o and the + 1st-order diffracted light D _{p on} the pupil plane Ep of the projection optical system PL in this case are shown in FIG.
(Details will be described later). The resolution limit of the pattern by the conventional illumination method is determined by whether or not ± 1st order diffracted light can pass through the projection optical system. That is, conventionally, when the pattern pitch is P, the pattern size given by P> λ / NA is the resolution limit. On the other hand, in the multi-inclined illumination method in this embodiment, almost P
> Λ / 2NA is the resolution limit.

Further, in FIG. 2, 0th-order diffracted light D _o and +
The first-order diffracted light D _p passes through an optical path that is substantially symmetrical with respect to the optical axis AX. This is because the incident angle ψ of the illumination light flux Lo is sin
It can be realized by setting ψ = sin θ / 2 = λ / 2P. At this time, when defocusing the wafer W, 0
The order diffracted light D _o and the + 1st order diffracted light D _p generate almost the same amount of wavefront aberration (due to defocus). This is because the wavefront aberration with respect to the defocus amount ΔF is ½ × ΔF sin ² t (where t is the incident angle of each diffracted light on the wafer), and here the 0th-order diffracted light and the + 1st-order diffracted light are incident on the wafer. This is because the incident angles t are almost equal. By the way, the cause of collapsing (defocusing) the pattern image on the wafer is the difference in wavefront aberration between the respective light beams. However, in the apparatus used in this embodiment (for example, FIG. 5), the 0th-order diffracted light D _o and the + 1st-order diffracted light D _p irradiated on the wafer have almost the same wavefront aberration, and therefore the defocus amount ΔF is the same. However, the degree of blurring is less than that of the conventional exposure apparatus. That is, the depth of focus is deep.

Further, in FIG. 2, 0th-order diffracted light D _o and +
The first-order diffracted light D _p is assumed to be substantially symmetrical with respect to the optical axis AX, but in order to increase the depth of focus, the pupil plane Ep
It is not necessary that these two diffracted lights are symmetrical with respect to the optical axis AX, and may pass through positions that are substantially equidistant from the optical axis AX. Here, since the fineness (line width, pitch) and directionality of the pattern of the reticle used in the exposure apparatus are not specified to one type, the pupil of the illumination optical system of the projection exposure apparatus used in this embodiment is not limited. The center position of the local area through which the illumination light flux passes on the surface, for example, the positions of the two fly-eye lens groups 7B ₁ and 7B ₂ in the pupil plane of the illumination optical system shown in FIG. 5 are variable according to the type of pattern. Is desirable. That is, this is because each of the incident direction and the incident angle of the illumination light flux on the reticle is determined by the direction in which the pattern is drawn, the width, and the pitch. For example, in the case of a one-dimensional line-and-space pattern arranged in the X direction as shown in FIG. 2, the illumination light flux L0 is
It may be incident from the direction shown in the figure. That is, the illumination light flux L0 is incident on the reticle R parallel to the paper surface. In addition, 0th-order diffracted light D ₀ and + generated from the pattern
If the incident angle of the illumination light flux L0 is determined so that the first-order diffracted light D _p passes through the position on the pupil plane Ep that is substantially equidistant from the optical axis AX, the depth of focus of the image on the wafer W can be increased. it can.

The eccentric direction from the optical axis AX of the central position of the local area where the illumination light flux passes within the pupil of the illumination optical system is determined according to the incident direction thus determined, and the line width of the pattern is determined. Or the optical axis A according to the incident angle determined by the periodicity
The amount of eccentricity from X will be determined. The center position of the local area through which the illumination light flux passes in the pupil plane of the illumination optical system is determined as described above.

When illuminating a pattern drawn in a two-dimensional direction, the direction of the illuminating light beam is adjusted to 2 in accordance with the direction of each pattern.
Illuminate from the direction. Further, the illuminating luminous fluxes are in two directions, that is, two luminous fluxes are added, and two luminous fluxes which are symmetric with respect to the optical axis of the projection optical system are added to the two luminous fluxes.
Use light flux. In that case, since the light amount centroid direction of these light fluxes coincides with the optical axis of the projection optical system, it is possible to prevent the lateral displacement (telecentric displacement) of the image that occurs when the wafer is slightly defocused. Also, FIG.
In the case of illuminating a reticle having a one-dimensional line-and-space pattern as shown in (1), it is also preferable to perform multi-tilt illumination with two light beams that are substantially symmetrical with respect to the optical axis of the projection optical system. When illuminating a one-dimensional pattern, if the intensity ratio of the 0th-order diffracted light and the 1st-order diffracted light generated from the pattern is exactly 1: 1, illumination is performed with two light beams from a direction symmetrical to the optical axis. No need. However, this is limited to the case where there is only one pattern pitch. Regarding the above
Next, a detailed description will be given with reference to FIGS. 3 and 4.

FIG. 3 shows an apparatus used in this embodiment (FIG. 1).
2 is a diagram schematically showing a part of FIG. 1, in which the components from the fly-eye lens group to the projection optical system PL in FIG. 1 are shown. still,
In FIG. 3, the relay lenses 9 and 11 and the variable blind 1 are shown.
0 is omitted. Now, the fly-eye lens group (here, the two fly-eye lens groups 7B ₁ and 7B shown in FIG.
₂ is arranged in the illumination optical path), the exit end of the reticle pattern substantially coincides with the Fourier transform plane (pupil plane of the illumination optical system) 70. Here, the focal length of the condenser lens 12 that brings the pupil plane 70 of the illumination optical system and the pattern plane of the reticle R into a Fourier transform relationship is set to f as shown in FIG.

4 (B) and 4 (D) are shown in FIG.
The position of the illumination light flux (local area) for maximizing the depth of focus for each of the reticle patterns RP ₁ and RP ₂ shown in (C) in the pupil plane 70 of the illumination optical system (hereinafter, simply referred to as a secondary This is referred to as the position of a light source group (surface light source). Further, as described above, the pupil plane 70 of the illumination optical system and the reticle pattern RP
_{Since 1} has a Fourier transform relationship, the position of the secondary light source group (corresponding to the exit end surface of the fly-eye lens group) in the pupil plane 70 of the illumination optical system is determined by the angle of incidence of the illumination light L0 on the reticle R. Corresponds to sin ψ in ψ (Fig. 2). At this time, since the focal length of the condenser lens 12 is f as shown in FIG. 3, the distance from the optical axis AX of the center positions of the fly-eye lens groups 7B ₁ and 7B _{2 on} the pupil plane 70 of the illumination optical system is f.
× sin ψ.

FIG. 4A shows a so-called one-dimensional line-and-space pattern, in which a transmissive portion Ra and a light-shielding portion R are formed.
b extend in the Y direction and are regularly arranged at a pitch P in the X direction. At this time, the individual fly-eye lens group 7
The optimum positions of B ₁ and 7B ₂ (that is, the positions of the secondary light source groups) are on the line segment Lα in the Y direction assumed in the pupil plane 70 of the illumination optical system as shown in FIG. It is an arbitrary position on Lβ. FIG. 4B is a view of the pupil plane 70 of the illumination optical system corresponding to the reticle pattern RP ₁ seen from the optical axis AX direction, and the coordinate system XY defined in the pupil plane 70 is the reticle pattern from the same direction. It is the same as FIG. 4A in which RP ₁ is viewed.

Now, as shown in FIG. 4B, in the pupil plane 70, the distances α and β from the optical axis AX (center of the pupil) to the line segments Lα and Lβ are α = β, and λ is the exposure wavelength. Then, α = β = f · (1/2) · (λ / P). As described above, if these distances α and β can be expressed as f × sin ψ, sin
ψ = λ / 2P, which is in agreement with the above-mentioned numerical value. Therefore, if the positions of the centers of the fly-eye lens groups 7B ₁ and 7B ₂ (that is, the centers of gravity of the light amount distribution of the secondary light source group) are on the line segments Lα and Lβ, the line and space shown in FIG. With respect to the pattern, the _two diffracted lights of the 0th-order diffracted light and the ± 1st-order diffracted light generated by the illumination light from each of the fly-eye lens groups 7B ₁ and 7B ₂ are generated by the projection optical system PL. Optical axis AX at pupil plane Ep
You will pass through a position that is almost equidistant from. Therefore,
As described above, the line and space pattern (Fig. 4
The depth of focus for (A) can be maximized, and high resolution can be obtained.

On the other hand, FIG. 4C shows the reticle pattern.
This is the case when the isolated space pattern is
Turn RP₂X direction (horizontal direction) pitch is Px, Y direction
It is assumed that the pitch (vertical direction) is Py. Figure 4
(D) is each fly-eye lens group 7B in this case₁, 7
B₂It is a figure showing the optimal position of, and the position with FIG. 4 (C),
The rotational relationship is the same as the relationship shown in FIGS. 4 (A) and 4 (B).
Two-dimensional pattern RP ₂In the case of, the pupil plane of the illumination optical system
At 70, four local regions (illumination luminous flux) are formed.
Therefore, in the device having the above configuration (FIG. 1),
It is sufficient to provide four fly-eye lens groups (see the figure
7D in 5₁~ 7D_Four).

Now, the two-dimensional pattern R as shown in FIG.
When the illumination light is incident on P ₂ , diffracted light is generated in the two-dimensional direction according to the periodicity (X: Px, Y: Py) of the pattern RP _{2 in} the two-dimensional direction. Also in the two-dimensional pattern RP ₂ as shown in FIG. 4C, one of the 0th order diffracted light and the ± 1st order diffracted light in the diffracted light is the pupil plane Ep of the projection optical system PL.
The depth of focus can be maximized by passing through a position that is substantially equidistant from the optical axis AX. Figure 4
In the pattern RP ₂ of (C), the pitch in the X direction is Px, so α = β = f · (1/2) as shown in FIG. 4 (D).
If the centers of the two fly-eye lens groups (secondary light source groups) are on the line segments Lα and Lβ that are (λ / Px), the depth of focus is maximized for the X-direction component of the two-dimensional pattern RP _2. be able to. Similarly, γ = ε = f · (1/2) ·
If the centers of the two fly-eye lens groups (secondary light source groups) are on the line segments Lγ and Lε that are (λ / Py), the depth of focus is maximized for the Y-direction component of the two-dimensional pattern RP _2. You can

As described above, the illumination light flux from the fly-eye lens group (secondary light source group) arranged at each position in the pupil plane 70 shown in FIG. 4 (B) or (D) becomes the reticle pattern RP ₁ or RP ₂ . Upon incidence, either the + 1st-order diffracted light component D _p or the −1st-order diffracted light component D _m from the reticle pattern and the 0th-order diffracted light component D _o are from the optical axis AX at the pupil plane Ep of the projection optical system PL. It will pass through positions that are almost equidistant. Therefore, as described above, a projection type exposure apparatus having a high resolution and a large depth of focus can be realized.

Here, the reticle pattern shown in FIG.
Although only the two examples shown in (A) or (C) were considered, the + 1st-order diffracted light component or -1 from the pattern is focused on the periodicity (fineness) of other patterns.
The pupil plane 70 of the illumination optical system is so arranged that two light fluxes of one of the second-order diffracted light components and the 0th-order diffracted light component pass through the position on the pupil plane Ep of the projection optical system PL that is substantially equidistant from the optical axis AX. Fly eye lens group (secondary light source group)
Set the center position of. The other diffracted light, for example, one of ± 2nd-order diffracted lights and the 0th-order diffracted light may have a positional relationship that is substantially equidistant from the light source AX on the pupil plane of the projection optical system.

Further, when the reticle pattern includes a two-dimensional periodic pattern as shown in FIG. 4D, one specific 0
Focusing on the second-order diffracted light component, the pupil plane E of the projection optical system PL
On p, a first-order or higher-order diffracted light component distributed in the X direction (first direction) centering on the one zero-order diffracted light component,
There may be first-order or higher-order diffracted light components distributed in the Y direction (second direction). Therefore, assuming that a two-dimensional pattern is satisfactorily formed on one specific 0th-order diffracted light component, one of the higher-order diffracted light components distributed in the first direction,
One of the higher-order diffracted light components distributed in the second direction and a specific 0
The three 0th-order diffracted light components are distributed in the pupil plane Ep at almost the same distance from the optical axis AX.
It suffices to adjust the position of the fly-eye lens group) as the two secondary light source groups. For example, in FIG. 4D, the respective center positions of the fly-eye lens group (secondary light source group) are set to points Pζ, Pη, P.
It is better to match either κ or Pμ. Points Pζ, P
η, Pκ, and Pμ are all line segments Lα or Lβ (the optimum position for the periodicity in the X direction, that is, the 0th-order diffracted light and X
Direction ± 1st-order diffracted light is located at an equal distance from the optical axis on the pupil plane Ep of the projection optical system), and the line segment Lγ or Lε (the optimal position for the periodicity in the Y direction). Therefore, the light source position is optimal in either the X direction or the Y direction.

In the above description, a pattern having a two-dimensional directivity at the same location on the reticle is assumed as a two-dimensional pattern. The above method can be applied to. When the pattern on the reticle has a plurality of directions or fineness, the optimum position of each fly-eye lens group corresponds to each direction and fineness of the pattern as described above. Alternatively, each fly-eye lens group may be arranged at the average position of each optimum position. Further, this average position may be a weighted average in which a weight corresponding to the fineness or importance of the pattern is added.

Further, the light fluxes emitted from the respective fly-eye lens groups are obliquely incident on the reticle. At this time, if the direction of the light amount center of gravity of these inclined incident light beams (plurality) is not perpendicular to the reticle, the position of the transferred image shifts in the in-plane direction of the wafer when the wafer W is slightly defocused. To do. In order to prevent this, the direction of the light amount centroid of the illumination light fluxes (plurality) from each fly-eye lens group is set to be perpendicular to the reticle pattern, that is, parallel to the optical axis AX. That is, when an optical axis (center line) is assumed for each fly-eye lens group, a position vector in the Fourier transform plane of the optical axis (center line) with reference to the optical axis AX of the projection optical system PL, and It suffices that the vector sum of the product of the light amount emitted from the fly-eye lens group becomes substantially zero. Also, as an easier way,
The number of secondary light sources is set to 2m (m is a natural number), and the positions of m of them are determined by the above-described optimization method (FIG. 4), and the remaining m are positions that are substantially symmetrical with respect to the m and the optical axis AX. You can place it in. Next, with reference to FIG. 5, a change mechanism of the fly-eye lens group used for changing the illumination condition, particularly for switching between the normal illumination and the plurality of inclined illuminations will be described. As shown in FIG. 5, in the present embodiment, four types of fly-eye lens groups 7A to 7D, which are fixed by changing the eccentricity with respect to the optical axis AX in accordance with the difference in the periodicity of the reticle pattern, are attached to the holding member 7. It is integrally provided, and by driving (rotating) the holding member 7 by the drive system 54, each of the plurality of fly-eye lens groups 7A to 7D can be replaceably arranged in the optical path of the illumination optical system. ing. Here, when each of the fly-eye lens groups 7B to 7D is arranged in the illumination optical path, the fly-eye lens groups are arranged substantially symmetrically with respect to the optical axis AX. still,
In FIG. 1, a fly-eye lens group 7 used for normal illumination
It is shown that A is arranged in the illumination optical path. In FIG. 5, for example, fly-eye lens groups 7B ₁ and 7B ₂
Are arranged in the illumination optical path, their centers (in other words, the centers of gravity of the respective light amount distributions formed by the secondary light source images in each of the fly-eye lens groups 7B ₁ and 7B ₂ ) depend on the periodicity of the reticle pattern. The fly-eye lens groups 7B ₁ and 7B ₂ are integrally held by a holding member (rotary turret plate) 7 so as to be set at discrete positions decentered with respect to the optical axis AX by a determined amount. In addition, fly eye lens group 7
In each of B to 7D, it is desirable that the position of each fly-eye lens group can be moved according to the periodicity of the reticle pattern.

Here, for example, the fly-eye lens group 7
It is desirable that each of B ₁ and 7B ₂ has the same shape and the same material (refractive index). Further, each lens element of the individual fly-eye lens group 7A shown in FIG. 1 is a biconvex lens, and the light source side focal plane coincides with the incident surface, and the reticle side focal plane coincides with the exit surface. However, the lens elements of the fly-eye lens group do not have to strictly satisfy this relationship, and the lens elements may be plano-convex lenses, convex-plano lenses, or plano-concave lenses.

FIG. 5 is a diagram showing a specific structure of the holding member, and here, four types of fly-eye lens groups 7 are used.
A to 7D are arranged at intervals of about 90 ° on a holding member (turret plate) 7 that is rotatable around the rotation shaft 7a. Each of the four types of fly-eye lens groups 7A to 7D is
The fly-eye lens groups hold different eccentric states (that is, positions in a plane substantially perpendicular to the optical axis AX) with respect to the optical axis AX (center of the holding member) depending on the periodicity of the reticle pattern. Has been done. The fly-eye lens groups 7B and 7C are both two fly-eye lens groups (7B
₁ , 7B ₂ ), (7C ₁ , 7C ₂ ), and when these fly-eye lens groups are arranged in the illumination optical system, they are fixed so that their arrangement directions are substantially orthogonal to each other. . Fly-eye lens unit 7D are approximately equidistant to place the four fly's eye lens 7D ₁ ～7D ₄ from its center 7d (optical axis AX), and fixed. Fly's eye lens group 7
A is the optical axis A when its center is placed in the illumination optical system.
It is fixed so that it almost coincides with X, and is used when the conventional exposure is performed.

As is apparent from FIG. 5, the turret plate 7 is rotated by the drive element 55 including a motor and a gear according to the information of the reticle bar code BC as described above, and the four types of fly-eye lens groups 7A to 7A. 7
Each D can be exchanged, and a desired holding member according to the periodicity (pitch, arrangement direction, etc.) of the reticle pattern can be arranged in the illumination optical system. At this time, the driving element 5
Since 5 is also used for positioning, it is desirable to provide a rotation angle measuring member such as a rotary encoder. In addition, each of the plurality of fly-eye lens groups,
As shown in FIG. 5, it is composed of 16 lens elements (36 elements only for the fly-eye lens group 7A), but it is not limited to this, and in the extreme case, a fly element composed of one lens element is used. It may be an eye lens group.

Further, four types of fly-eye lens groups 7A to
In each of 7D, a portion other than the fly-eye lens group is used as a light-shielding portion within a circle (corresponding to the irradiation area of the illumination light flux) shown in FIG. 5 or a light-shielding member is provided behind the turret plate 7 (on the reticle side). It is desirable to arrange. Further, when it is necessary to strictly set the incident angle of the illumination light flux to the reticle pattern, each of the plurality of fly-eye lens groups in the holding member is set in the radial direction (radiation direction) about the optical axis AX. It may be configured such that it can be slightly moved, and each of the plurality of fly-eye lens groups can be rotated around the optical axis AX.

Further, as is apparent from FIG. 1, in this embodiment, a lens system having a large diameter (condensing light) is incident on each of the four types of fly-eye lens groups 7A to 7B so that the illumination light beam enters the flood. Although the optical system 6) was used, the loss of the amount of illumination light that irradiates the reticle R becomes large in such a configuration. Therefore, in order to prevent light amount loss, an optical member (input optical system) for concentrating the light amount distribution on the light source side focal plane of the fly's eye lens group only near each fly eye lens position as shown in FIG. It is desirable to provide it.

FIG. 6 is a view showing a modification of the holding member. The input optical system (optical fiber bundles 71, 72) and the holding members (82, 84) are integrally movable members (support rods 8).
It is fixed to 6). Although the case where an optical fiber bundle is used is described here, the input optical system may be another member, for example, a diffraction grating or a polygonal prism.

In FIG. 6, the two fly-eye lens groups 80A and 80B are integrally held by a holding member 82,
The optical fiber bundle 71 has an entrance 71a and an exit 71b.
Both are held by the fixture 83, and the holding member 82 is integrally fixed to the fixture 83. Further, the inside of the holding member 82 has a fly-eye lens group 80A, 80B.
It is a light-shielding portion (hatched portion in the figure) except for. on the other hand,
The replacement fly-eye lens groups 81A and 81B are integrally held by a holding member 84, the optical fiber bundle 72 is held by a fixture 85 at both the incident portion 72a and the emission portion 72b thereof, and the holding member 84 is also a fixture. It is integrally fixed to 85, and like the above, the inside thereof serves as a light shielding portion. Further, the fixtures 83 and 85 are connected and fixed by a connecting member 87. Therefore, when replacing the holding member, it is sufficient to replace the fixing member together. In FIG. 6, the fixture 83 is present in the illumination optical system, and the replacement fixture 85 is used.
Is set at a position outside the illumination optical system. Also,
The light source side of the condensing optical system 6 and the reticle side of the relay lens 9 are assumed to have the same configuration as in FIG. 1, for example.

By the way, the replacement of the holding member is performed by the driving element 8
8 by pushing and pulling the support rod 86. Therefore, as shown in FIG. 6, when the fly-eye lens group is replaced, the fly-eye lens group and the optical fiber bundle can be integrally replaced, so that the integrated member group (fixture) and the entire illumination optical system. There is an advantage that it is not necessary to adjust the positions of the respective members (the fly-eye lens group, the optical fiber bundle, etc.) for each replacement only by aligning the positions. At this time, since the drive element 88 is also used for positioning, it is desirable to provide a position measuring member such as a linearly encoder or a potentiometer.

The fly-eye lens group for each holding member shown in FIGS. 5 and 6 and the number of lens elements forming the fly-eye lens group may be arbitrary, and the fly-eye lens group and the incident surface of the lens element may be arbitrary. Alternatively, the shape of the exit surface is not limited to a rectangle. Also, FIGS.
Each position (position in the plane perpendicular to the optical axis) of the plurality of fly-eye lens groups shown in, that is, the holding member to be selected is determined (changed) according to the reticle pattern to be transferred.
Good to do. As described above, the determination (selection) method in this case is to obtain the optimum resolution and the effect of improving the depth of focus with respect to the fineness (pitch) of the pattern to be transferred by the illumination light flux from each fly-eye lens group. As described above, a holding member having a fly-eye lens group at or near the position of incidence on the reticle pattern (incident angle ψ) may be used.

When an optical fiber bundle (FIG. 6) is used as an optical member (input optical system) for concentrating the illumination light flux near each of the plurality of fly-eye lens groups, the fly-eye lens group Each of the plurality of emission ends is moved along with the movement, for example, the emission end and the fly-eye lens group may be integrally fixed. Further, although the rectangular fly's eye lens group also tilts relatively with the rotation of the holding member, when the holding member is rotated, only the position of the fly eye lens group moves without causing the tilt. It is desirable to configure When the fly-eye lens group is replaced, the variable aperture stop 8 arranged near the input optical system or its exit surface is used.
Since it is also necessary to replace them, it is desirable to integrally fix these optical members to the holding member 7 for each fly-eye lens group.

In each of the plurality of fly-eye lens groups 7A to 7D shown in FIG. 5, the exit end area of each fly-eye lens group is the numerical aperture of the illuminating light flux passing through the fly-eye lens group with respect to the reticle R. It is desirable that the ratio of the numerical aperture (NA _R ) of the projection optical system PL to the reticle side, that is, the so-called σ value is set to about 0.1 to 0.3. When the σ value is smaller than 0.1, the pattern fidelity of the transferred image is deteriorated, and when it is larger than 0.3, the effect of improving the resolution and increasing the depth of focus is weakened. Also, the condition of the σ value determined by one of the fly-eye lens groups (0.1 ≦ σ ≦
In order to satisfy the above condition, the size of the exit end area of each fly-eye lens group (the size in the in-plane direction perpendicular to the optical axis) is determined according to the illumination light flux (emission light flux). Is also good.

A variable aperture stop may be provided in the vicinity of the reticle-side focal plane of each fly-eye lens group to change the numerical aperture of the light flux from each fly-eye lens group to change the σ value. At the same time, in cooperation with the variable aperture stop (NA limiting stop) 32 arranged near the pupil plane of the projection optical system PL, the NA of the projection system and the σ value can be further optimized.

The light flux entering each fly-eye lens group is illuminated to a degree outside the entrance end surface of each fly-eye lens group to a certain extent, and the distribution of the amount of light entering each fly-eye lens group is uniform. It is preferable because the illuminance uniformity on the reticle pattern surface can be further enhanced.
For example, in FIG. 6, another fly array lens may be added to the light source side of the input optical systems 71 and 72, that is, in the vicinity of the incident surfaces of the incident portions 71a and 72a to form a two-stage integrator structure. At this time, the first-stage (light source side) fly-eye lens is a single one, that is, one large fly-eye lens that spreads over the entire pupil plane of the illumination optical system. The reticle side focal plane (emission side focal plane) of the first-stage (light source side) fly-eye lens is arranged in the plane in the vicinity of the pupil plane of the illumination optical system. further,
The exit surface of the first-stage fly-eye lens is a Fourier transform surface with respect to the entrance surface of the second-stage fly-eye lens group. However, it is not necessary to strictly maintain the relationship of the Fourier transform so that the light flux emitted from each element of the first-stage fly-eye lens is superposed on the incident surface of the second-stage fly-eye lens. It is good if the relationship is maintained. Further, although the fly-eye type integrator is used as the optical integrator in the present embodiment, for example, a prismatic rod-type integrator may be used.
Further, a prismatic rod type integrator and a fly's eye type integrator may be combined to form the two-stage integrator structure as described above. Such a structure is disclosed in Japanese Patent Laid-Open No. 1-271718.

Further, in this embodiment, when the illumination condition is changed corresponding to the reticle pattern, the turret plate 7 is rotated to arrange the desired fly-eye lens group in the illumination optical path. The fly-eye lens group to be placed near the pupil plane of the system is only one large fly-eye lens that spreads over the entire pupil plane, and it corresponds to the fineness of the reticle pattern within the pupil plane of the illumination optical system or its conjugate plane. A plurality of apertures (for example, the fly-eye lens group 7) so as to limit the illumination light flux passing through to at least one local region having a center at a position decentered from the optical axis of the illumination optical system.
A plurality of spatial filters provided with B ₁ and 7B ₂ ) may be exchangeably arranged and arranged in the vicinity of the exit surface (reticle side focal plane) of the fly-eye lens group.
At this time, it is desirable to make each element of the fly-eye lens group fine (to reduce the cross-sectional area) in consideration of the illuminance uniformity on the reticle surface.

Next, an example of changing the illumination condition and the light amount distribution in the vicinity of the pupil plane of the projection optical system under each illumination condition will be described with reference to FIG. In FIG. 8, members having the same functions and functions as those in FIG. 1 are designated by the same reference numerals. Now, the coherence factor of the illumination optical system, the so-called σ value, is
Numerical aperture NA _IL of illumination optical system and numerical aperture NA _{R of} projection optical system
Ratio to (reticle side), that is, σ = NA _IL / NA _R =
It is represented by sin θ ₁ / sin θ ₂ . Here, as shown in FIG. 8A, θ ₁ is the angle of the illumination light that irradiates the reticle R,
θ ₂ is the angle of the illumination light that can pass through the aperture stop 32 of the projection optical system PL. FIG. 8 shows a case where the numerical aperture NA _R of the projection optical system is kept constant and the numerical aperture NA _IL of the illumination optical system is changed, and FIG. 8A shows the case where the σ value is large and FIG.
Indicates when the σ value is small. As is clear from FIG. 8, in this embodiment, the σ value can be arbitrarily set by the variable aperture stop 8 arranged near the exit surface of the fly-eye lens group 7A (that is, the pupil plane of the illumination optical system).
In the present embodiment (FIGS. 1 and 8), the variable aperture stop is used as a means for changing the σ value. Even if the illumination optical system (fly-eye lens group or the like) is configured to be replaceable, the σ value can be made variable in the same manner as above.

As shown in FIGS. 8A and 8B, the 0th-order diffracted light of the illumination light flux that has passed through the reticle R enters the projection optical system PL at the same angle θ ₁ as the incident angle of the illumination light. ,
As shown by the solid line in the drawing, the wafer W is reached. Although not shown, the first-order and higher-order diffracted light travels outside the 0th-order diffracted light and reaches the wafer W. A variable aperture stop 32 is provided on the pupil plane Ep of the projection optical system PL, and the numerical aperture NA _W (wafer side) of the projection optical system PL is NA _W = sin θ ₃
It is represented by. Here, looking at the light quantity distribution on the pupil plane of the projection optical system PL, the 0th-order diffracted light passes through almost the center (hatched portion) of the pupil plane within the plane defined by the aperture stop 32, and the first-order or higher order The diffracted light will pass through the outside. Generally 1
Since the intensity of the 0th-order diffracted light is higher than that of the diffracted light of the higher order,
Especially for the lens elements near the pupil plane, it is considered that the illumination light flux is concentrated at the center of the pupil and the temperature rises when the σ value is small (FIG. 8B) and larger (FIG. 8A). To be When pattern exposure is actually performed, the optimum illumination conditions corresponding to these types of reticle (for example, normal reticle or phase shift reticle), pattern line width, shape, periodicity, etc. Numerical aperture NA
_W (further, the fly-eye lens group) is set by driving the variable aperture diaphragms 8 and 32 (turret plate 7).

Next, the operation of the projection exposure apparatus according to this embodiment will be described. First, a method of calculating the amount of change in the image forming characteristic of the projection optical system PL will be briefly described. In the present embodiment, four types of image forming characteristics, that is, projection magnification, field curvature, distortion, and focus position are dealt with, but these image forming characteristics have different generation mechanisms and their variation characteristics are different. Each imaging characteristic needs to be calculated independently as they may differ from each other. However, since the calculation methods of these imaging characteristics are the same, the projection magnification will be described as an example here. Further, in the present embodiment, the case where the σ value of the illumination optical system is changed as the illumination condition will be described.

Generally, the temperature of an object when heat is absorbed by the object is determined by the balance between the amount of heat absorbed by the object and the amount of heat released from the object. This is generally called a first-order lag system, and it is considered that the change characteristic of the projection magnification has the same behavior. Fig. 9 shows the fluctuation characteristics of the first-order lag system.
Shown in. FIG. 9 shows the change characteristic of the projection magnification when the projection optical system PL is irradiated with a certain amount of illumination light for a certain period of time. The change characteristic shown in FIG. 9 can be determined by two values, that is, the ratio ΔM / E of the final amount ΔM of change of the projection magnification with respect to the irradiation energy, and the time constant T indicating the change over time. Figure 9
, The time constant T is ΔM with respect to the final change amount ΔM.
It can be expressed as a time that changes by x (1-e ^-1 ). That is, the smaller the time constant T, the faster the change. The ratio ΔM / E and the time constant T may be obtained by measuring the projection magnification while actually irradiating the projection optical system PL with the illumination light to obtain the change characteristic as shown in FIG. 9. Actually, since the structure of the projection optical system PL is complicated, it does not show a simple change characteristic as shown in FIG. 9 and may be expressed as the sum of several first-order delays. In the example, in order to simplify the explanation,
A case where a simple change characteristic is shown will be described.

Now, in FIG. 1, when the reticle R is replaced, the wafer stage WS is driven to move the photoelectric sensor 33 to the optical axis position of the projection optical system PL, and the amount of illumination light incident on the projection optical system PL is measured. . Next, main controller 50 determines ratio ΔM / E obtained in advance, time constant T, irradiation energy detected by photoelectric sensor 33, and shutter 3.
The amount of change in the projection magnification is calculated by calculation based on the opening / closing time.

Here, when the σ value of the illumination optical system changes as described above, the light quantity distribution in the lens element near the pupil plane Ep of the projection optical system PL, that is, the temperature distribution changes. For example, when the σ value is small, the irradiation energy is concentrated in the center of the pupil and the temperature of the central part becomes high (FIG. 8 (B)).
It is considered that the lens element undergoes large thermal deformation near the center. On the other hand, when the σ value is large (FIG. 8 (A)),
It is considered that the temperature of the lens element as a whole rises and a particularly large thermal deformation cannot occur. Therefore, even if the total irradiation energy amount is the same, the change of the projection magnification is larger when the σ value is smaller than when the σ value is large,
That is, the ratio ΔM / E is considered to be large.

On the other hand, when the .sigma. Value is small, the temperature of the peripheral portion does not rise so much, and it is considered that the lens temperature suddenly drops when the irradiation of the illumination light is interrupted. On the other hand, when the σ value is large, the temperature of the lens gradually decreases because the lens temperature is increasing as a whole. Therefore, it is considered that when the σ value is small, the temporal change is larger than when the σ value is large, that is, the time constant T is small. From the above, it can be seen that both ΔM / E and T are changed by changing the σ value.

Therefore, next, the σ value is changed and the exposure is continuously performed.
Think about when you do. In addition, it is obtained at this time
FIG. 10 shows how the change amount ΔM of the projection magnification is. First, OK
Drive the variable aperture stop 8 to set a small σ value of the illumination optical system.
After that, time t₁Starts exposure from. At this time, the projection
The change amount ΔM of the magnification is the curve K shown by the solid line in FIG.₁To
Change along. Actually, the movement of the wafer stage
Illumination light is projected during (during stepping) and wafer exchange
The amount of change ΔM decreases without entering the optical system PL, and the curve K₁
Shows a jagged curve, but on average
Growls Therefore, as exposure continues, the change amount ΔM
Saturation level M₁Reach Where time t₂Exposure motion at
If the operation is stopped, the amount of change ΔM is the curve K indicated by the dotted line. ₂Along
Will decrease.

Next, time t₃(≈t₂) With variable aperture stop
Same as when σ value is small by driving 8 to increase σ value.
When the irradiation energy enters the projection optical system PL,
Consider the case of resuming light. Where time t
₃Then, if ΔM = 0, the change amount ΔM is shown in FIG.
Curve K shown as a line₃Change along with. But Naga
In the projection optical system PL, the condition before the illumination condition change (σ value is
Since the history under (when small) remains,
For example, change K₁I think it will look like (solid line)
Curve K₂And K₃Cannot be expressed as the sum of This
This is time t₃~ T _FiveBetween, near the pupil plane of the projection optical system
The temperature distribution (light amount distribution) of the lens element is shown in FIG.
Not in either state (A) or (B), and the lens
This is because the temperature cannot be expressed as a simple sum. Therefore, the curve K
₁In order to accurately determine the
The temperature distribution in the state is calculated and the corresponding change amount Δ
We have to know M, but it is not
It is theoretically possible, but in reality it is very complex and practical
There is no.

An exposure operation in the embodiment of the present invention for solving the above problem will be described with reference to FIG. Also in FIG. 11, the variable aperture stop 8 is used as in FIG.
After the σ value is set to a small value by (FIG. 8 (B)), time t
Open the shutter 3 at ₁ and start exposure. As shown in FIG. 11, the change amount ΔM of the projection magnification is equal to the curve K _{1 in} FIG.
Similar to (solid line), it changes along the curve indicated by the solid line, and eventually reaches the saturation level M ₁ .

At time t ₂ , the shutter 3 is closed to change the illumination condition and the exposure operation is stopped, and then the variable aperture diaphragm 8 is driven to set a large σ value (FIG. 8A).
At this time, main controller 50 stops the exposure operation at time t ₂ and then continues for a predetermined period of time, for example, level M where variation ΔM is constant.
_The exposure operation continues to be stopped until it decreases to _S1 . If the exposure operation continues to be stopped as it is, the amount of change ΔM decreases along the curve L ₁ (dotted line) and becomes zero at time t ₄ . Here, the level M _{S1 which} is the criterion (threshold value) for starting the exposure operation
The method of determining will be described later.

It should be noted that from time t _{1 the} change amount ΔM changes to the level M _S1.
Until the time t _{3 at} which the value reaches the time t ₃ , the change amount ΔM is the calculation parameter (ΔM / E described above, when the σ value is small,
It is calculated by calculation using T). Further, during the pattern exposure (time t _{1 to} t ₂ ), the image formation characteristic is constantly set to a predetermined value by driving the image formation characteristic correction mechanism based on the calculated value.

If the change amount ΔM reaches a constant level M _S1 at time t ₃ , the main controller 50 sets the calculation parameters ΔM / E and T to the calculation parameters when the σ value is large. After the switching (correction), the shutter 3 is opened and the exposure operation is restarted. At this time, the variation amount ΔM of the projection magnification is calculated by using the previously corrected calculation parameter, and the pattern exposure of the wafer W is executed while the projection magnification is corrected by the imaging characteristic correction mechanism. .
At this time, the time (t ₃ −t ₂ ) has already elapsed since the end of the exposure operation when the σ value is set small.
In other words, since the change amount ΔM has decreased to the level M _S1 , the influence of the history when the σ value with respect to the change amount ΔM is small is sufficiently small. Therefore, even if the change amount ΔM actually changes, for example, as indicated by a curved line L ₂ shown by the dotted line in FIG. 11, the calculation error is small, and the imaging characteristics required for the projection exposure apparatus. Is negligible with respect to the control accuracy of.

[0081] Incidentally, Continuing exposure while setting a large σ value, the variation ΔM at time t ₅ reaches the saturation level M _2, have stopped the exposure operation at time t ₅ in Figure 11. Further, as described above, when the σ value is large, the temperature of the lens element rises as a whole, so the change amount ΔM after time t ₅ is gentle as compared with the change amount ΔM from time t _{2 to} t _3. Will be reduced to.

In the above description, the σ value is set small, and the illumination condition (σ value) is changed when the change amount ΔM reaches the saturation level M _S1 . The σ value may be changed before ΔM reaches the saturation level M ₁ . For example, in FIG. 11, when the exposure operation is stopped at time t ₂ ′ and the σ value is set large, the exposure operation is continuously stopped until time t ₃ ′ when the variation ΔM decreases to the level M _S1. . This state,
It is shown as a curve L ₄ (dotted line) in FIG. 11. In this case, the stop time (t ₃ '-t _2') above the stop time (t ₃ -
It is shorter than t ₂ ).

Next, a method of determining the constant level M _S1 used for determining the exposure start will be described. Basically, the level M _S1 is the calculation parameter ΔM /
Even if E and T are switched, the difference (calculation error) between the image formation characteristic (change amount ΔM) calculated under the calculation parameter and the actual image formation characteristic (corresponding to the curve L ₂ in FIG. 11). May be determined to be within a predetermined allowable value (uniquely determined according to the control accuracy of the imaging characteristic). At this time, it is desirable to set the level M _{S1 to a} value as large as possible so as not to reduce the throughput of the apparatus, that is, to shorten the time (t ₃ −t ₂ ).

For example, when the change of the σ value is minute, or when the pattern to be transferred does not require strict control accuracy regarding the projection magnification, the level M _S1 is set to a relatively large value and the In this case, the level M _S1 must be set small and the exposure operation must be stopped until the influence of the history on the imaging characteristics under the illumination condition before the change is sufficiently small. Actually, the above relationship was previously tested,
Alternatively, it is desirable to determine the value of the level M _S1 by obtaining it by simulation or the like and storing it in the form of a table or an approximate expression inside the main controller 50. It should be noted that, in the attenuation calculation of the change amount ΔM after time t ₂ in FIG. 11, it is not necessary to perform a special calculation because the change amount ΔM has been obtained by successive calculation in the conventional manner.

Next, an example of the processing in the main controller 50 will be described with reference to the flowchart shown in FIG.
Although not shown in FIG. 1, the main control unit 50 controls the overall system including the variable aperture diaphragms 8 and 32, and a calculation for calculating the amount of change in the imaging characteristics as described above. It is assumed that it is composed of In FIG. 12, when the σ value is not changed as a normal operation, the calculation unit calculates the change amount ΔM of the current projection magnification of the projection optical system in step 102, and the control unit calculates the change amount ΔM in step 103 based on the calculation result. A predetermined drive command is output to the image formation characteristic correction mechanism to correct the amount of change ΔM. Then, unless the arithmetic unit receives the signal that the σ value has been changed in step 101 from the keyboard 51, the bar code reader 52, or the program (exposure sequence) input in advance, the control unit repeatedly executes the above operation. To do. That is, the arithmetic unit is in the loop, that is, in steps 101 to 1 surrounded by a dotted line in FIG.
Only during 03, the control unit continues to output a signal that exposure is possible. At this time, the control unit continues to perform the exposure operation on the wafer while correcting the projection magnification by the image formation characteristic correction mechanism.

On the other hand, if it is determined in step 101 that the σ value has been changed, the arithmetic unit stops outputting a signal that exposure is possible to the control unit, and step 104
Proceed to. In step 104, the arithmetic unit determines the level M _S1 which is the criterion for restarting the exposure operation based on the change width of the σ value and the allowable control accuracy (allowable magnification error) of the projection magnification, and controls this level M _S1 . Set to the department. In the next step 105, the calculation unit calculates the change amount ΔM of the projection magnification of the current projection optical system (however, the value calculated under the calculation parameter before changing the illumination condition), and the calculated value and the level M. Compare the size with _S1 . If it is determined here that the calculated value ΔM is larger than the level M _S1 , in step 107 the arithmetic unit warns the operator that the exposure operation needs to be stopped. Is displayed. Thereafter, in step 108, the arithmetic unit further causes the control unit to perform a predetermined time Δt (where Δt is, for example, the time (t ₃ −t ₂ ) shown in FIG. 11).
It is set shorter), and the output of the command to restart the exposure operation is stopped. Then, the process returns to step 105 again to obtain the change amount ΔM (calculated value) of the projection magnification, and in step 106, the magnitude of the calculated value ΔM and the level M _S1 is determined. Hereinafter, steps 105 to 108 are repeatedly executed until the calculated value ΔM becomes smaller than the level M _S1 , and the calculated value ΔM
_Becomes smaller than the level M _S1 , the process proceeds to step 109 to erase (cancel) the warning on the display device. Further, in step 110, the calculation unit sets the calculation parameter (ΔM / E, T) corresponding to the new illumination condition (σ value),
Returning to step 101, the normal exposure operation (steps 101 to 103) is repeatedly executed under the new calculation parameter. As a result, it becomes possible to perform the pattern exposure on the wafer always under the highly accurate imaging characteristics.

As described above, in the present embodiment, only the projection magnification is described as the imaging characteristic of the projection optical system, but other aberrations can be calculated and corrected in the same manner, and the projection optical system can be connected. It is possible to always maintain the image characteristics with high accuracy. Further, when the same calculation and correction are performed for a plurality of aberrations, when determining the stop time of the exposure operation, the change amount of all the aberrations to be corrected is equal to or less than the reference value (corresponding to the level M _S1 ). The exposure operation may be restarted at the point.

Further, in the above embodiment, the case where the σ value of the illumination optical system is changed as the illumination condition has been described, but the present invention can be applied to obtain the same effect when changing other illumination conditions. it can. For example, switching from normal illumination to multiple-tilt illumination (or annular illumination) (or vice versa), that is, fly eye lens group 7A to fly eye lens group 7B
7D to 7D (FIG. 5), the variable blind 10 is driven to change the illumination field of view when the fly-eye lens groups 7B to 7D are exchanged in accordance with the periodicity of the reticle pattern even in multi-tilt illumination. The present invention is also effective for changing, switching from normal illumination to annular illumination (or vice versa), or changing the numerical aperture of the projection optical system PL. When performing multi-tilt illumination, replace the fly-eye lens group and set the σ value to 0.1
It is desirable to set it to about 0.3. FIG. 13 (A),
FIG. 13B shows the light quantity distributions in the vicinity of the pupil plane of the projection optical system when annular illumination and multiple tilt illumination are performed, respectively, and FIG.
(C) shows the light quantity distribution in the vicinity of the pupil plane of the projection optical system when the numerical aperture of the projection optical system is changed. FIG. 13 (C)
Then, compared with FIGS. 13A and 13B, the variable aperture stop 3
The opening of 2 is large.

By the way, when the numerical aperture of the projection optical system is changed by the variable aperture stop 32, the light quantity distribution becomes as shown in FIG. 13C, and the distribution of the 0th order diffracted light does not change. Therefore, it is considered that the above-mentioned change has a small influence on the image forming characteristic, but the distribution of the diffracted light of the first order or higher can change and affect the control accuracy of the image forming characteristic. It is desirable to apply the present invention even when making changes. In this case, the illumination condition of the illumination optical system is not changed, but the illumination condition change in the present invention includes the change of the numerical aperture of the projection optical system.

Further, even when the ordinary reticle is replaced with the phase shift reticle (or vice versa), the same effect can be obtained by applying the present invention. For example, when a spatial frequency modulation type phase shift reticle (disclosed in Japanese Patent Publication No. 62-50811) is used, the light amount distribution as shown in FIG. . When using a phase shift reticle, it is desirable to set the σ value to about 0.1 to 0.3 as an illumination condition.

Further, in FIG. 1, when the reticle R is loaded on the reticle stage RS, the bar code reader 5
Based on the reticle information (name) read in 2, the reticle type (normal reticle or phase shift reticle),
Alternatively, an example has been shown in which the illumination conditions most suitable for the fineness of the pattern (periodicity, pitch, direction, etc.) are set for each reticle, but the illumination conditions may be changed for the same reticle.

By the way, in the explanation of FIG. 10 of the above embodiment,
And the curve K₁Is the curve K₂And curve K ₃Expressed as the sum of and
However, the imaging characteristics are not so precise.
When throughput is important and there is no need to control
Is approximately K₁= K₂+ K₃Changes in imaging characteristics
A method of calculating the quantity may be adopted. In this case, beforehand
In what lighting conditions, reticle or its pattern,
To find out how much the calculation error of the imaging characteristics occurs
Then, after comparing with this error tolerance,
It should be done only when possible.

Further, as shown in FIG. 7, by using the reference member 15, the image forming characteristics of the projection optical system (focal position, field curvature, etc.).
The throughput is usually measured once a day, for example, about once a day because the throughput decreases when the measurement is performed frequently. However, after the illumination condition is changed as in the above embodiment, the exposure operation is stopped for a certain period of time. May employ a method of continuing the exposure operation while frequently performing the above measurement. In this case, although there is a decrease in throughput due to the above measurement, the decrease in throughput due to the stop of the exposure operation can be reduced, and the decrease in overall throughput can be reduced.

Further, in the above embodiment, not only the exposure operation is stopped, but also the projection optical system is actively adapted to the new illumination condition. Therefore, the exposure is not actually performed, and the projection optical system is changed to the new illumination condition. It is also possible to apply the method of irradiating under the light. In this case, for example, a shutter is arranged between the projection optical system PL and the wafer W to prevent the illumination light from reaching the wafer, or to prevent the wafer W from being exposed during the irradiation. It is necessary to retract the wafer stage WS. Further, when measuring the image forming characteristics using the reference member 15 while the exposure operation is stopped, a spatial filter that satisfies the same new illumination condition is placed near the exit surface of the optical fiber 62. It is desirable to perform the above measurement by

[0095]

As described above, according to the present invention, even when the change characteristic of the image forming characteristic due to the absorption of the illumination light of the projection optical system changes with the change of the illumination condition, the change of the image forming characteristic before the change is performed. After stopping the exposure operation until the influence of the history under the illumination condition becomes small, the exposure operation is started under the new illumination condition while controlling the imaging characteristics. For this reason, it is possible to prevent the calculation of the change amount of the image formation characteristic due to the absorption of the illumination light from being difficult because the change characteristics of the image formation characteristic under the two illumination conditions before and after the change of the illumination condition are mixed. The amount of change in the image forming characteristic can be obtained exactly in the pattern exposure with respect to, and the image forming characteristic can be corrected with high accuracy.

Further, since the stop time of the exposure operation is determined according to the control accuracy of the image forming characteristic corresponding to the fineness of the mask pattern, it is possible to minimize the decrease in productivity (throughput) due to the stop time. It is possible.

[Brief description of drawings]

FIG. 1 is a diagram showing a schematic configuration of a projection exposure apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating the principle of an example of an illumination method (multi-slope illumination) adopted in the embodiment of the present invention.

FIG. 3 is a diagram schematically showing an optical path from a fly-eye lens group to a projection optical system.

FIG. 4 is a diagram illustrating an arrangement of fly-eye lens groups in multiple tilt illumination.

FIG. 5 is a diagram showing a specific configuration of an exchange mechanism that exchangeably arranges a plurality of fly-eye lens groups in an optical path of an illumination light beam.

FIG. 6 is a diagram showing an example of an input optical system that focuses an illumination light flux on each of a plurality of fly-eye lens groups.

FIG. 7 is a diagram for explaining an operation of measuring the image formation characteristic of the projection optical system using the reference member.

FIG. 8 is a diagram showing an optical path from a fly-eye lens group to a wafer and a light amount distribution near a pupil plane of a projection optical system before and after changing an illumination condition (σ value).

FIG. 9 is a diagram for explaining a change characteristic of an image forming characteristic (projection magnification) of a projection optical system due to absorption of illumination light.

FIG. 10 is a diagram showing how the projection magnification changes when the exposure is continuously performed while changing the σ value.

FIG. 11 is a diagram for explaining an exposure operation of the projection exposure apparatus according to the embodiment of the present invention.

FIG. 12 is a flowchart showing an example of a control operation of the image forming characteristic of the projection optical system.

FIG. 13 is a diagram showing a light quantity distribution in the vicinity of a pupil plane of the projection optical system when the numerical aperture of the projection optical system is changed and the annular illumination, the multiple tilt illumination are used.

[Explanation of symbols]

 3 Shutter 7A-7D Fly-eye lens group 8, 32 Variable aperture diaphragm 10 Variable blind 15 Reference member 20-23 Lens element 25, 27, 29 Drive element 30, 31 Focus detection system 33 Irradiation amount monitor (photoelectric sensor) 50 Main control Device 51 Keyboard 52 Bar Code Reader 53 Drive Element Control Unit 54 Drive System R Reticle RS Reticle Stage PL Projection Optical System Ep Pupillary Surface W Wafer WS Wafer Stage BC Bar Code

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁵ Identification code Office reference number FI technical display location 7352-4M H01L 21/30 311 S

Claims (3)

[Claims] 1. An illumination optical system for shaping illumination light from a light source into a substantially uniform intensity distribution and irradiating the mask having a fine pattern with the uniform illumination light, and an image of the pattern of the mask on a photosensitive substrate. A projection exposure apparatus comprising a projection optical system for image-projecting and a means for controlling an imaging characteristic of the projection optical system, wherein an illumination condition is changed to change an illumination condition for the mask corresponding to the mask or its pattern. A projection exposure apparatus comprising: a means and an exposure control means for stopping the exposure operation on the photosensitive substrate for a predetermined time when the illumination condition is changed. 2. The change of the illumination condition is a change of a distribution state of the illumination light on a Fourier transform plane of the pattern of the mask of the projection optical system or a plane in the vicinity thereof. 1. The device according to 1. 3. The apparatus according to claim 1, wherein the change of the illumination condition is a change of the pattern configuration of the mask.