JP4185102B2 - Exposure equipment - Google Patents

Description

  The present invention relates to an exposure apparatus.

  With the demand for miniaturization of semiconductor integrated circuits, the exposure wavelength of the exposure apparatus is also required to be shortened. At present, ArF excimer laser (193 nm), F2 laser (157 nm) and the like have entered the practical stage. Yes. When using such short-wavelength exposure light, calcium fluoride crystals (fluorite) should be used as the material for the optical members (lenses, etc.) that make up the optical system in consideration of good transmittance. Is increasing. However, fluorite has a large birefringence, and has a drawback that wavefront aberration is generated depending on the polarization direction of incident light, or the wavefront is split to reduce the coherence of light. For this reason, in order to design an optical system with small aberrations, it is necessary to numerically grasp the magnitude of birefringence of the optical system.

  In order to indicate the magnitude of birefringence of a planar optical member, it is generally performed to identify the birefringence direction (azimuth angle) θ and the phase difference Δ. This method is not suitable for expressing the birefringence of an optical system having a plurality of optical members (lenses and the like) having a curved surface. For this reason, Jones matrix and Mueller matrix are generally used to show the characteristics of an optical system having birefringence.

The Jones matrix is a 2-by-2 matrix that represents the optical characteristics of an optical element that passes completely polarized light that can be expressed by a Jones vector. The Jones vector expresses the electric field vector (E) = (E _x , E _y ) by the amplitude a _x , a _y and the phase difference δ as follows.

  When perfectly polarized light that can be expressed by such a Jones vector passes through the optical element and undergoes some conversion, the Jones matrix represents the characteristics of the optical element. For example, a Jones matrix of a polarizer with an azimuth angle θ = 0 ° is expressed as follows.

  The Jones matrix can handle only completely polarized light (light that does not include an unpolarized component), and cannot handle partially polarized light that includes an unpolarized component. For this reason, the Mueller matrix that can handle the non-polarized component is suitable for expressing the birefringence characteristics of the optical system including the fluorite lens in which generation of the non-polarized component is unavoidable. The Mueller matrix is a 4 × 4 matrix that represents the optical characteristics of the optical element through which the partially polarized light expressed by the Stokes parameter passes.

The Stokes parameters are as follows, where s ₀ is the total intensity of light, s ₁ is the intensity of the 0 ° linearly polarized component, s ₂ is the intensity of the 45 ° linearly polarized component, and s ₃ is the intensity of the clockwise circularly polarized component. In particular, for completely polarized light, it is expressed as follows.

In this Stokes parameter, (s ₀ ² − (s ₁ ² + s ₂ ² + s ₃ ² )) ^1/2 represents a non-polarized component. That is, according to the Stokes parameter, partially polarized light including a non-polarized component can be expressed.

When partially polarized light expressed by such a Stokes parameter passes through the optical element and undergoes some conversion, the Mueller matrix represents the characteristics of the optical element. For example, the Mueller matrix (P ₀ ) of a polarizer with an azimuth angle θ = 0 ° is expressed as follows. In the following description, the Mueller matrix of the azimuth angle θ is represented as a symbol (P _θ ).

  If this Mueller matrix can be measured, it will be possible to easily calculate how the optical element converts the polarization state for light of any polarization. As an apparatus for measuring a Mueller matrix with respect to a single optical member, for example, a Mueller ellipsometer described in Patent Document 1 is known.

  However, in the Mueller ellipsometer described in Patent Document 1 and the like, a single optical member is a measurement target, and an optical system composed of a plurality of optical members already incorporated in an exposure apparatus or the like is a measurement target. It is not something to be done.

In recent years, in an exposure apparatus, for the purpose of increasing the contrast of a pattern to be projected, exposure light having a specific polarization direction is projected. However, if the birefringence and non-polarization components generated in the projection optical system cannot be ignored, the magnitude of this birefringence and the non-polarization components must be quantitatively evaluated, and the polarization direction of the exposure light taking this into account must be selected. The polarization direction of the exposure light may not be as desired. Further, the characteristics of the projection optical system and the like of the exposure apparatus can change due to secular change, dirt adhesion, and the like. Therefore, it has been desired to measure the Mueller matrix such as the projection optical system of the exposure apparatus without decomposing the exposure apparatus.
JP 2000-502461 A (Pages 19-22, FIGS. 1-4)

  The present invention has been made in view of the above points, and an object thereof is to provide an exposure apparatus that can measure the Mueller matrix of an optical system included in the apparatus without disassembling the exposure apparatus.

  An exposure apparatus according to an aspect of the present invention includes a light source that emits a light beam, an illumination optical system that guides the light beam to a photomask, and a projection optical system that projects an image of a mask pattern onto the object to be exposed. In the exposure apparatus, the mask pattern by the projection optical system, the first polarization conversion apparatus capable of sequentially converting the light beam into any one of a plurality of polarized lights having a perpendicular relationship on the Poincare sphere and outputting them. A light intensity detector that is mounted in the vicinity of the imaging position of the image and detects the light intensity of the converted polarized light after the polarized light has passed through the illumination optical system, the photomask, and the projection optical system; A second polarization conversion device that includes a linear polarizer and a linear phase shifter and is inserted in the optical path of the converted polarized light; and switches between the azimuth angle of the linear polarizer and the insertion / removal of the linear phase shifter or the azimuth angle. By the light intensity detector To obtain a plurality of data of the light intensity detected, characterized in that a parameter calculator for calculating said transformation Stokes parameters of polarized light based on the data of a plurality of light intensity obtained.

At this time, the linear polarizer can be set to have an azimuth angle of 45 × n ° (n is an integer). In addition, the parameter calculation unit separates the linear phase shifter or makes the linear polarizer and the azimuth coincide with each other and sets the azimuth of the linear polarizer to 0 °, 90 °, 45 °, and −45 °. Then, data on the light intensity detected by the light intensity detector at each azimuth angle is acquired, and the linear phase shifter is inserted at a predetermined azimuth angle different from the azimuth angle of the linear polarizer and the linearly polarized light The azimuth angle of the child is switched between 0 ° and 90 °, or between 45 ° and −45 °, and the light intensity data detected by the light intensity detector before and after this switching is acquired, and these The Stokes parameter of the converted polarized light can be calculated based on the light intensity data. Alternatively, the parameter calculation unit may remove the linear phase shifter or make the linear polarizer and the azimuth angle coincide with each other and set the azimuth angle of the linear polarizer to 0 ° and 90 °. The light intensity data detected by the light intensity detector in FIG. 3 is obtained, and the Stokes parameters (s ₀ ′, s ₁ ′, s ₂ ′, s _{3) of the} converted polarized light are obtained based on the light intensity data. ) S ₀ ′ and s ₁ ′ of “)” are calculated, and the linear phase shifter is separated or the azimuth angle of the linear polarizer is matched and the azimuth angle of the linear polarizer is set to 45 ° and −45 °. The light intensity data detected by the light intensity detector at each azimuth angle is set and the Stokes parameters (s ₀ ′, s ₁ ′) of the converted polarized light are obtained based on the light intensity data. , S ₂ ′, s ₃ ′) s ₀ 'And s ₂ ' are calculated, the linear phase shifter is inserted at a predetermined azimuth angle different from the azimuth angle of the linear polarizer, and the azimuth angle of the linear polarizer is between 0 ° and 90 °, or 45 The light intensity data detected by the light intensity detection unit is acquired before and after the switching, and the Stokes parameter of the converted polarized light is obtained based on the light intensity data. It may be configured to calculate s ₀ ′ and s ₃ ′ of (s ₀ ′, s ₁ ′, s ₂ ′, s ₃ ′).

  According to the exposure apparatus of the present invention, the Mueller matrix of the optical system incorporated in the exposure apparatus can be measured without disassembling the apparatus, whereby the polarization direction of the projection light was measured. It is possible to select according to the Mueller matrix, which can contribute to the improvement of the performance of the exposure apparatus.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  First Embodiment FIG. 1 shows a configuration of an exposure apparatus 100 to which a first embodiment of the present invention is applied. As shown in FIG. 1, the exposure apparatus 100 includes a light source device 1, a first polarization conversion device 2, an illumination optical system 3, a photomask 5, a projection optical system 6, a second polarization conversion device 7, a photodetector 8, An arithmetic control unit 11, a driver 12, a driver 13, and an A / D converter 14 are provided. In FIG. 1, reference numeral 10 denotes an image plane on which an object to be exposed such as a wafer is placed. The light source device 1 is, for example, a laser light source device that emits exposure light having a predetermined wavelength and polarized in a predetermined direction. Here, the light source device 1 emits exposure light that is linearly polarized in the horizontal direction, that is, 0 °. And Since the illumination optical system 3, the photomask 5, and the projection optical system 6 are the same as known ones, detailed description thereof is omitted.

  The first polarization conversion device 2 and the second polarization conversion device 7 have a function of changing the polarization state of the light beam passing therethrough in order to measure the Mueller matrix of the projection optical system 6 and the like. Accordingly, the first polarization conversion device 2 and the second polarization conversion device 7 are inserted into the optical system of the exposure apparatus 100 only at the time of measuring the Mueller matrix such as the projection optical system 6, and at normal times (for example, during the exposure operation). Leave the light path. Instead of separating the light beam from the optical path, it is possible to set the change in the polarization state of the light beam passing through to be zero.

The first polarization conversion device 2 is for converting the polarization direction of the exposure light emitted from the light source device 1, and as shown in FIG. 2, a half-wave plate 21 and a quarter-wave plate 22. The half-wave plate 21 is held so as to be rotatable around the optical axis O of the illumination optical system 3, whereby the azimuth angle can be changed in increments of 22.5 degrees. Instead of the half-wave plate 21, a combination of a depolarizing plate and a linear polarizer capable of changing the azimuth angle may be used. Similarly, the quarter-wave plate 22 is held rotatably around the optical axis O of the illumination optical system 3, so that the azimuth angle can be changed in increments of 45 degrees. Specifically, when the azimuth angle of the half-wave plate 21 and the azimuth angle of the quarter-wave plate 22 are set as shown on the left side of the following [Table 1], it passes through the first polarization conversion device 2. The polarization direction of the polarized light (S ⁽ⁱ⁾ ) (i = 1 to 6) can be converted as shown on the right side of [Table 1]. The quarter-wave plate 22 is inserted so that the azimuth is 45 ° with respect to the polarization direction of the light beam that has passed through the linear polarizer 21 when circularly polarized light is emitted from the first polarization conversion device 2. In other cases, it is inserted such that the polarization direction and the azimuth angle of the light beam that has passed through the linear polarizer 21 coincide. Note that the rotation of the half-wave plate 21 and the quarter-wave plate 22 is controlled by the driver 12 based on a control signal from the arithmetic control unit 11. Further, the half-wave plate 21 and the quarter-wave plate 22 are arranged in a square shape slightly inclined with respect to the optical axis in order to prevent generation of return light to the light source device 1.

That is, in this embodiment, the polarized light (S ⁽¹⁾ ), (S ⁽²⁾ ), (S ⁽³⁾ ), (S ⁽⁴⁾ ), (S ^{( 5)} ) and (S ⁽⁶⁾ ) are represented by six intersections of the pointe sphere PB and the coordinate axes X, Y, and Z on the pointe sphere PB as shown in FIG. These polarized lights (S ⁽¹⁾ ), (S ⁽²⁾ ), (S ⁽³ ), (S ⁽⁴⁾ ), (S ⁽⁵⁾ ), (S ⁽⁶⁾ ) are expressed by Stokes vectors, It is expressed as the following [Equation 5].

However, polarized light (S ⁽¹⁾ ), (S ⁽²⁾ ), (S ⁽³⁾ ), (S ⁽⁴⁾ ), (S ⁽⁵⁾ ), (S ⁽⁶⁾ ) are not necessarily shown in FIG. There is no need to be a combination as shown in FIG. 2, and any combination of light that is orthogonal to each other on the Poincare sphere PB and light that is symmetrical with the origin may be used. For example, elliptically polarized light may be used.

  The second polarization conversion device 7 includes a quarter wavelength plate 71 and a linear polarizer 72 as shown in FIG. The quarter wavelength plate 71 has an azimuth angle fixed at −45 ° and is detachably held in the optical path as shown in FIG. Further, the linear polarizer 72 is rotatably held around the optical axis O, and the azimuth angle can be changed in increments of 45 °. Further, the second polarization conversion device 7 itself can be inserted into and removed from the optical path of the exposure apparatus 100, and the insertion position can be changed according to the position of the optical system or the like to be measured. The driver 13 performs the insertion / removal control of the quarter-wave plate 71 and the rotation control of the linear polarizer 72 based on a control signal from the arithmetic control unit 11.

Here, when it is assumed that the second polarization conversion device 7 is not provided, the above polarized light (S ⁽¹⁾ ), (S ⁽²⁾ ), (S ⁽³⁾ ), (S ⁽⁴⁾ ), Polarized light (converted polarized light) after (S ⁽⁵⁾ ) and (S ⁽⁶⁾ ) are converted by passing through the projection optical system 6 is converted into (S ⁽¹⁾ ′) and (S ⁽² ), respectively. ⁾ '), (S ⁽³⁾ '), (S ⁽⁴⁾ '), (S ⁽⁵⁾ ') and (S ⁽⁶⁾ ').

In the present embodiment, conversion polarized light (S ⁽¹⁾ ′), (S ^{( 1)} is performed by inserting and removing the quarter wavelength plate 71 and changing the azimuth angle of the linear polarizer 72 as described later. ²⁾ The light receiving state on the photodetector 8 of '), (S ⁽³⁾ '), (S ⁽⁴⁾ '), (S ⁽⁵⁾ ') and (S ⁽⁶⁾ ') is changed. As shown in FIG. 1, the photodetector 8 is disposed in the vicinity of the imaging plane 10 that is the position where the image of the mask pattern 5 is formed by the projection optical system 6. Based on the change in the light receiving state, the converted polarized light (S ⁽¹⁾ ′), (S ⁽²⁾ ′), (S ⁽³⁾ ′), (S ⁽⁴⁾ ′), (S ^{( 5)} The polarization states of ') and (S ⁽⁶⁾ ') are detected as Stokes vectors. The detection signal of the photodetector 8 is input to the arithmetic control unit 11 by the A / D converter 14. The calculation control unit 11 calculates the Stokes parameter of the converted polarized light (S ′) based on this detection signal and the control states of the first polarization conversion device 2 and the second polarization conversion device 7 at that time, and also projects the projection. The Mueller matrix of the optical system 6 etc. is determined.

  The relationship between the polarized light (S) and the converted polarized light (S ′) can be expressed by the Mueller matrix (M) indicating the optical characteristics of the projection optical system 6 as follows.

Therefore, polarized light (S ⁽¹⁾ '), (S ⁽²⁾ '), (S ⁽³⁾ '), (S ⁽⁴⁾ '), (S ⁽⁵⁾ ') and (S ⁽⁶⁾ ' ) Is detected as a Stokes vector, the elements of each column of the Mueller matrix (M) can be calculated as the following [Equation 7] to [Equation 10].

  Next, in the present embodiment, a procedure for measuring the Mueller matrix (M) of the projection optical system 6 will be described.

First, the azimuth angles of the half-wave plate 21 and the quarter-wave plate 22 of the first polarization conversion device 2 are adjusted to emit polarized light (S ⁽ⁱ⁾ ) (i = 1 to 6), and Mueller After passing through the projection optical system 6 in which the matrix (M) is unknown and converted into converted polarized light (S ⁽ⁱ⁾ '), the light is received by the photodetector 8. As shown in FIG. 1, the second polarization conversion device 7 is disposed immediately after the projection optical system 6 that is a measurement target.

In this state, first, the quarter-wave plate 71 of the second polarization conversion device 7 is detached from the optical path, the azimuth angles of the linear polarizer 72 are set to 0 ° and 90 °, and the photodetectors at the respective azimuth angles. 8 output signals are obtained. The Mueller matrix (P ₀ ) of a linear polarizer with an azimuth angle of 0 ° and the Mueller matrix (P ₉₀ ) of a linear polarizer with an azimuth angle of 90 ° are expressed by the above [Equation 4] and the following [Equation 11], respectively. Expressed.

Therefore, the Stokes parameters of the polarized light (P ₀ S ′) and (P ₉₀ S ′) on the photodetector 8 when the azimuth angle of the linear polarizer 72 is set to 0 ° and 90 ° are as follows: [Expression 12] and [Expression 13].

  At this time, the light intensities measured by the photodetector 8 are as shown in the following [Equation 14] and [Equation 15], respectively.

Therefore, s ₀ ′ and s ₁ ′ can be determined as follows.

Similarly, for s ₂ ′ and s ₀ ′, the quarter wavelength plate 71 of the second polarization conversion device 7 is separated from the optical path, and the azimuth angle of the linear polarizer 71 is 45 ° and −45 ° (or 135 °). And can be calculated by obtaining the output signal of the photodetector 8 at each azimuth angle. Note that the Mueller matrix (P ₄₅ ) of a linear polarizer with an azimuth angle of 45 ° and the Mueller matrix (P ₋₄₅ ) of a linear polarizer with an azimuth angle of −45 ° are the following [Equation 18] and the following [Equation 19], respectively. ] Is expressed as follows.

Accordingly, s ₂ ′ and s ₀ ′ can be obtained by the following equations by the same method as that for obtaining s ₁ ′ and s ₀ ′.

Next, in order to measure s ₃ ′, in the second polarization conversion device 7, a quarter wavelength plate 71 having an azimuth angle of −45 ° is inserted, and the azimuth angles of the linear polarizer 71 are set to 0 ° and 90 °. The light intensity of the light received by the photodetector 8 at each azimuth angle is measured.

Here, the Mueller matrix (Q _θ ) of the quarter wavelength plate of the azimuth angle θ is

It is expressed as Therefore, the exposure light (P ₀₎ generated when the exposure light (S ′) passes through the ¼ wavelength plate 71 having an azimuth angle of −45 ° and the linear polarizer 72 having an azimuth angle of 0 ° or 90 °. The Stokes parameters of (Q ₋₄₅ S ′) and (P ₉₀ Q ₋₄₅ S ′) are expressed as the following [Equation 23] and [Equation 24].

s ₃ ′ and s ₀ ′ are determined as follows.

Thus, when the Stokes parameter elements s ₀ ′ to s ₃ ′ of the converted polarized light (S ′) are obtained, the Mueller matrix (M) of the projection optical system 6 is obtained by the above [Equation 7] to [Equation 10]. The elements of each column are obtained.

  In the first embodiment, the first polarization converter 2 uses the half-wave plate 21 and the quarter-wave plate 22, and the half-wave plate 21 is the optical axis O of the illumination optical system 3. The emitted light from the light source device 1 is converted into a desired polarization state by changing the azimuth angle around the 22.5 degree around. Instead, the first polarization conversion device 2 has an azimuth angle of 45. You may be comprised with the optical rotator and the quarter wavelength plate which can be changed in steps.

  In the first embodiment, the quarter-wave plate 71 is used in the second polarization conversion device 7, but this is only an example, and in general, a phase difference between two orthogonal electric vectors is obtained. Linear phase shifters that give Δ can be used.

In the first embodiment, in order to measure s ₃ ′, a quarter wavelength plate 71 having an azimuth angle of −45 ° is inserted in the second polarization conversion device 7 and the azimuth of the linear polarizer 71 is also measured. Although the angle is switched between 0 ° and 90 °, the present invention is not limited to this. For example, a quarter wavelength plate 71 having an azimuth angle of 90 ° may be inserted, and the azimuth angle of the linear polarizer 71 may be switched between 45 ° and −45 °.

  In the first embodiment, the quarter wavelength plate 71 is separated from the optical path when it is not necessary. However, the linear polarizer 72 and the azimuth angle may be made to coincide with each other instead of being separated.

  Further, instead of switching the azimuth angle of the linear polarizer 72 and inserting / removing the quarter wavelength plate 71 or switching the azimuth angle, the azimuth angle of the linear polarizer 72 is fixed at 0 °, for example, and the quarter wavelength plate 71 The light intensity may be measured while changing the azimuth angle.

[Second Embodiment]
Next, a second embodiment of the present invention will be described.

  In the first embodiment, the Mueller matrix of the projection optical system 6 along the optical axis O is measured. On the other hand, in the second embodiment, light is blown in various directions in the pupil of the projection optical system 6 using a photomask 5 having a diffraction grating pattern whose period and direction are changed. The distribution of the Mueller matrix in the pupil of the system 5 is measured.

For this reason, the converted polarized light (S ′) is also incident on the quarter-wave plate 71 obliquely. In general, in a quarter wave plate, the retardation changes according to the incident angle of incident light. For this reason, when the converted polarized light (S ′) is obliquely incident on the quarter-wave plate 71, the quarter-wave plate 71 needs to be handled as a general linear polarizer. For this reason, in the present embodiment, when measuring s ₃ ′ among the Stokes parameter elements of the converted polarized light (S ′), the quarter wavelength plate 71 is subjected to two conditions of azimuth angles of −45 ° and 45 °. Measurement is performed, and at each azimuth angle, the azimuth angle of the linear polarizer 71 is further switched between 0 ° and 90 °. Thus, 1/4 wavelength plate 71 Rita - even Deshon is changed by the incident angle of the light, can be measured s 3 _'with high precision.

For the measurement of the Stokes parameter elements s ₀ ′, s ₀ ′, and s ₂ ′ of the converted polarized light (S ′), the azimuth angle of the linear polarizer 72 is appropriately changed as in the first embodiment. Can be measured.

The measurement procedure of s ₃ ′ in the present embodiment will be specifically described.

First, the azimuth angle of the quarter wave plate 71 is set to 45 ° and inserted into the optical path, and in this state, the azimuth angle of the linear polarizer 72 is set to 0 °. At this time, the Mueller matrix of the quarter-wave plate 72 when the converted polarized light (S ′) is incident on the quarter-wave plate 72 whose azimuth angle is set to 45 ° from an oblique direction is represented by (R _Δ45 ) (Rita _−dance Δ), the light (P ₀ R _Δ45 S ′) passing through the second polarization conversion device 7 is

The light intensity I (P ₀ R _Δ45 S ′) detected by the photodetector 8 can be expressed as the following [ _Equation 28].

Similarly, when the azimuth angle of the quarter-wave plate 71 is set to 45 ° and the azimuth angle of the linear polarizer 72 is set to 90 °, light that passes through the second polarization conversion device 7 and is detected by the photodetector 8. I (P ₉₀ R _Δ45 S ′) can be expressed as the following [ _Equation 29].

Further, the ¼ wavelength plate 71 is set to an azimuth angle of −45 °, and the azimuth angle of the linear polarizer 71 is switched between 0 ° and 90 °, and the following [Equation 30] and [Equation 31]. Light having the expressed light intensity is detected by the photodetector 8. However, (R _Δ−45 ) is the wavelength of the quarter-wave plate 71 when the converted polarized light (S ′) is incident on the quarter-wave plate 72 whose azimuth is set to −45 ° from an oblique direction. Mueller matrix.

Therefore, for example, from [Equation 30] obtained by subtracting [Equation 29] and s ₁ ′ measured in advance by the same method as in the first embodiment, a quarter-wave plate 71 actual retardances Δ are determined.

Furthermore, s ₃ ′ can be determined based on [Equation 33] obtained by subtracting [Equation 31] from [Equation 29] and the obtained retardance Δ. Thus, the measurement result as shown in FIG. 6 can be expected by considering the incident angle dependence of the retardation of the quarter wavelength plate. In FIG. 6, the circle indicates the numerical aperture NA of the projection optical system 6.

[Third Embodiment]
Next, a third embodiment of the present invention will be described with reference to FIG.

In the first and second embodiments, the Mueller matrix of the projection optical system 6 is measured on the assumption that the photomask 5 has no polarization characteristic. However, the photomask may have birefringence due to a process such as polishing, or stress birefringence may occur when the photomask is fixed to the exposure apparatus. Therefore, in the present embodiment, the photomask 5 having such polarization characteristics is regarded as a linear phase shifter, and its Mueller matrix (R _Δθ ) is calculated separately from the projection optical system 6.

In order to calculate the Mueller matrix (R _Δθ ) of the photomask 5, in this embodiment, as shown in FIG. 7, the linear polarizer 15 is placed downstream of the photomask 5, specifically, the photomask 5 and the projection optics. Place between systems 6. The linear polarizer 15 can be inserted at two azimuth angles. Here, the azimuth angle can be set to two types of 0 ° and 45 °. At each azimuth angle of 0 ° and 45 °, the first polarization converter 2 gives two kinds of polarized light (here, clockwise circularly polarized light (S ⁽³⁾ ), counterclockwise circularly polarized light (S ⁽⁶⁾ )) From the four combinations, the light detector 8 detects four light intensities, and calculates the Mueller matrix (R _Δθ ) based on the detection result. The second polarization conversion device 7 is preferably detached from the optical path, but may be inserted in the optical path.

A specific procedure for calculating the Mueller matrix (R _Δθ ) of the photomask 5 will be described next.

First, the clockwise circularly polarized light (S ⁽³⁾ ) is emitted from the first polarization conversion device 2, and the azimuth angle of the linear polarizer 15 is set to 0 °. If the Mueller matrix of the projection optical system 6 is (M) (in this case, it means a Mueller matrix of only the projection optical system 6 that does not include the photomask 5 having polarization characteristics), the polarization that reaches the light detection unit 8 The converted light (MP ₀ R _Δθ S ⁽³⁾ ) is expressed by the following [Equation 34], and the light intensities I ₀ and _CW detected by the light detection unit 8 are expressed by the following [Equation 35]. .

Similarly, the light intensity I detected by the light detection unit 8 when the left-handed circularly polarized light (S ⁽⁶⁾ ) is emitted from the first polarization conversion device 2 and the azimuth angle of the linear polarizer 15 is set to 0 °. _{0 and CCW} are expressed by the following [ _Equation 36].

Similarly, the light intensity I detected by the light detection unit 8 when the clockwise polarized light (S ⁽³⁾ ) is emitted from the first polarization conversion device 2 and the azimuth angle of the linear polarizer 15 is set to 45 °. _{45, CW} is expressed by the following [Equation 37].

Similarly, the light intensity I detected by the light detection unit 8 when the left-handed circularly polarized light (S ⁽⁶⁾ ) is emitted from the first polarization conversion device 2 and the azimuth angle of the linear polarizer 15 is set to 45 °. _{45, CCW} is expressed by the following [ _Equation 38].

  From [Equation 35] to [Equation 38], the azimuth angle θ and retardance Δ when the photomask 5 is regarded as a linear phase shifter are independent of the Mueller matrix (M) of the projection optical system 6 alone. The following [Equation 39] and [Equation 40] are determined.

  The sign (+, −) of the retardance Δ shown in [Equation 40] cannot be determined by calculation alone. However, since the sign of the stress birefringence generated when the photomask 5 is clamped is sufficiently predictable, the sign can be easily determined from comparison with these predictions.

From the thus determined azimuth angle θ and retardance Δ, the Mueller matrix (R _Δθ ) of the photomask 5 can be determined.

  When the photomask 5 has polarization characteristics, the Mueller matrix (M ′) measured by the method shown in the first and second embodiments is the Mueller matrix (M ′) of only the actual projection optical system 6. ) Has the following relationship.

Thus, the actual projection optical system 6 only the Mueller matrix (M) can be the inverse matrix using the (R _{[Delta] [theta]} ^-1), as follows decisions (R _{[Delta] [theta]).}

[Fourth Embodiment]
Next, a fourth embodiment of the present invention will be described with reference to FIG. In the present embodiment, when not only the photomask 5 but also the illumination optical system 3 has polarization characteristics, it is intended to measure the Mueller matrix of the projection optical system 6 with high accuracy.

  In this embodiment, the illumination optical system 3 and the photomask 5 are regarded as one optical element, and its Mueller matrix is (R).

  When the illumination optical system 3 can be regarded as a linear phase shifter like the photomask 5 in the third embodiment, (R) is

It can be said. This is because an arbitrary R _Δθ can be expressed in this form no matter how many times. If the total Mueller matrix (R) of the illumination optical system 3 and the photomask 5 can be determined, the Mueller matrix (M) of only the projection optical system 6 can be determined as in the third embodiment.

  As shown in FIG. 8, a third polarization conversion device 9 composed of a combination of a linear polarizer 15 ′ and a quarter wavelength plate 16 is disposed between the photomask 5 and the projection optical system 6. Since the third polarization conversion device 9 has the same configuration as that of the second polarization conversion device 7, the second polarization conversion device 7 is disposed between the image mask 10 and the photomask 5 and the projection optical system 6. And may be used as a third polarization conversion device. Of course, the second polarization conversion device 7 can be left as it is, and a third polarization conversion device can be newly provided.

  The linear polarizer 15 ′ of the third polarization conversion device 9 can be switched between two azimuth angles between 0 ° and 90 °, and the quarter-wave plate 16 is detached from the optical path. The state (or the state in which the azimuth angle coincides with that of the linear polarizer 15 ′) and the state inserted in the optical path at an azimuth angle of −45 ° can be switched.

In the present embodiment, in the state where the azimuth angle of the linear polarizer 15 ′ is set to 0 ° and the quarter wavelength plate 16 is separated from the optical path, the first polarization conversion device 2 outputs 0 ° linearly polarized light (S ^{( 1)} ) and 90 ° linearly polarized light (S ⁽⁴⁾ ) are emitted to calculate the element r ₁₁ of the Mueller matrix (R).

First, when 0 ° linearly polarized light (S ⁽¹⁾ ) is projected, the converted polarized light reaching the photodetector 8 is assumed that the Mueller matrix of only the projection optical system 6 is (M).

And the light intensity of the converted polarized light detected by the photodetector 8 is

It is expressed. Next, when only the polarized light output from the first polarization conversion device 2 is changed from 0 ° linearly polarized light (S ⁽¹⁾ ) to 90 ° linearly polarized light (S ⁽⁴⁾ ), it is detected by the photodetector 8. The light intensity of the converted polarized light is

It is expressed. The [number 45] can be calculated as from the [number _{46], r 11} of the following [Expression 47].

Similarly, r ₁₂ is 45 ° linearly polarized light (S) from the first polarization conversion device 2 in a state where the azimuth angle of the linear polarizer 15 ′ is set to 0 ° and the quarter-wave plate 16 is separated from the optical path. By sequentially emitting ⁽²⁾ ) and −45 ° linearly polarized light (S ⁽⁵⁾ ), the following [Equation 48] can be calculated.

Similarly, r ₁₃ is a clockwise circularly polarized light (S) from the first polarization conversion device 2 in a state where the azimuth angle of the linear polarizer 15 ′ is set to 0 ° and the quarter wavelength plate 16 is separated from the optical path. By sequentially emitting ⁽³⁾ ) and counterclockwise circularly polarized light (S ⁽⁶⁾ ), the following [Equation 49] can be calculated.

In addition, r ₂₁ sets the azimuth angle of the linear polarizer 15 ′ to 45 ° and removes the 0 ° linearly polarized light (S ⁽ By sequentially emitting ¹⁾ ) and 90 ° linearly polarized light (S ⁽⁴⁾ ), the following [Equation 50] can be calculated.

R ₂₂ is 45 ° linearly polarized light (S ⁽²⁾ from the first polarization conversion device 2 in a state where the azimuth angle of the linear polarizer 15 ′ is set to 45 ° and the quarter-wave plate 16 is separated from the optical path. ⁾ ) And −45 ° linearly polarized light (S ⁽⁵⁾ ) are sequentially emitted, and can be calculated as the following [Equation 51].

Similarly, r ₂₃ is the clockwise circularly polarized light (S) from the first polarization conversion device 2 in a state where the azimuth angle of the linear polarizer 15 ′ is set to 45 ° and the quarter wave plate 16 is separated from the optical path. By sequentially emitting ⁽³⁾ ) and counterclockwise circularly polarized light (S ⁽⁶⁾ ), the following [Equation 52] can be calculated.

On the other hand, when calculating r ₃₁ , r ₃₂ , and r ₃₃ , the quarter-wave plate 16 is inserted into the optical path at an azimuth angle of −45 °.

First, r ₃₁ is set to 0 ° from the first polarization conversion device 2 in a state where the azimuth angle of the linear polarizer 15 ′ is set to 0 ° and the quarter wavelength plate 16 is inserted into the optical path at an azimuth angle of −45 °. The calculation can be performed by sequentially emitting linearly polarized light (S ⁽¹⁾ ) and 90 ° linearly polarized light (S ⁽⁶⁾ ).

When the 0 ° linearly polarized light (S ⁽¹⁾ ) is emitted, the converted polarized light reaching the photodetector 8 is

And the light intensity of the converted polarized light detected by the photodetector 8 is

It is expressed. When the polarized light output from the first polarization conversion device 2 is changed to 90 ° linearly polarized light (S ⁽⁴⁾ ), the light intensity of the converted polarized light detected by the photodetector 8 is

It becomes. The [number 54] can be calculated as from the [number _{55], r 31} of the following [Expression 56].

Similarly, r ₃₂ is set from the first polarization conversion device 2 in a state where the azimuth angle of the linear polarizer 15 ′ is set to 0 ° and the quarter wavelength plate 16 is inserted into the optical path at an azimuth angle of −45 °. By sequentially emitting 45 ° linearly polarized light (S ⁽²⁾ ) and −45 ° linearly polarized light (S ⁽⁵⁾ ), it is possible to calculate as in the following [Equation 57].

Similarly, r ₃₃ is set from the first polarization conversion device 2 in a state where the azimuth angle of the linear polarizer 15 ′ is set to 0 ° and the quarter wavelength plate 16 is inserted into the optical path at an azimuth angle of −45 °. By sequentially emitting clockwise circularly polarized light (S ⁽³⁾ ) and counterclockwise circularly polarized light (S ⁽⁶⁾ ), calculation can be performed as in the following [Equation 58].

  In this way, since the Mueller matrix (R) when the illumination optical system 3 and the photomask 5 are regarded as one optical system can be determined, correction using the inverse matrix of (R) makes it possible to perform projection optics. The Mueller matrix (M) of only the system 6 can be determined by the following [Equation 59].

  In addition, about the 3rd polarization converter 9, the change similar to the various change regarding the above-mentioned 2nd polarization converter 7 can be applied. For example, instead of the quarter-wave plate 16, a linear phase shifter that provides a phase difference Δ can be generally used. Further, instead of switching the azimuth angle of the linear polarizer 15 ′ and inserting / removing the quarter wavelength plate 16 or switching the azimuth angle, the azimuth angle of the linear polarizer 15 ′ is fixed at, for example, 0 °, and ¼ wavelength. The light intensity may be measured while changing the azimuth angle of the plate 16.

[Fifth Embodiment]
Next, a fifth embodiment of the present invention will be described. In this embodiment,
When the illumination optical system 3 and the photomask 5 are viewed as one optical element, it is an arbitrary polarizing element, that is, a Mueller matrix (R),

The case represented by is described. For example, it is assumed that an optical element such as a depolarization plate is incorporated in the illumination optical system 3. In [Equation 60], the element r ₀₀ at the upper left of the matrix (R) is set to 1. r ₀₀ is the intensity of the emitted light when illuminating unpolarized light, i.e., corresponds to the transmittance of the optical system, all the other elements in order that can be represented by factor multiplication of r _00.

In the present embodiment, in the configuration of the exposure apparatus shown in FIG. 8, the azimuth angle of the linear polarizer 15 ′ is switched between two (for example, 0 ° and 90 °, or 45 ° and −45 °, etc.) Two types of polarized light (eg, (S ⁽¹⁾ ) and (S ⁽⁴⁾ ) or (S ⁽²⁾ ) and (S ^{( 5)} ), or (S ⁽³⁾ ) and (S ⁽⁶⁾ ) etc.) are made incident on the illumination optical system 3, thereby calculating the elements of the Mueller matrix (R) shown in [Equation 60].

First, a method for calculating the elements r ₁₁ , r ₁₂ , and r ₁₃ in (R) will be described. The azimuth angle of the linear polarizer 15 ′ is set to either 0 ° or 90 °, and polarized light (S ⁽¹⁾ ) and (S ⁽⁴⁾ ) are sequentially transmitted from the first polarization conversion device 2 at each azimuth angle. Let it emit.

First, the following [Equation 61] is obtained by the combination of the azimuth angle 0 ° of the linear polarizer 15 ′ and the azimuth angle of the polarized light (S ⁽¹⁾ ).

Further, the following [Equation 62] is obtained by a combination of the azimuth angle 0 ° of the linear polarizer 15 ′ and the azimuth angle of the polarized light (S ⁽⁴⁾ ).

  Taking the ratio of the sum and difference of [Equation 61] and [Equation 62],

  It becomes. This ratio is

far. Similarly, the following [Equation 65] is obtained by setting the azimuth angle of the linear polarizer 15 ′ to 90 ° and projecting the polarized light (S ⁽¹⁾ ) and (S ⁽⁴⁾ ).

  The other illumination conditions are similarly expressed by the following [Equation 66] to [Equation 69].

Therefore, if r ₁₀ can be determined, r ₁₁ , r ₁₂ , and r ₁₃ are calculated as in the following [Equation 70] to [Equation 72] using the above [Equation 66] to [Equation 69]. Can do.

In order to obtain r ₂₁ , r ₂₂ , and r ₂₃ , the azimuth angles of the linear polarizer 15 ′ are set to 45 ° and −45 °, and the respective azimuth angles are symmetrical with respect to the origin on the Poincare sphere PB. A kind of polarized light is emitted. Thus, to obtain the following equation [73] - [Expression 78], further, in the [Expression 73] - [Expression _78], by determining the _{r 20, r 21, r 22} , r 23 following [Equation 79] to [Equation 81].

In order to obtain r ₃₁ , r ₃₂ , and r ₃₃ , the ¼ wavelength plate 16 is inserted into the optical path with an azimuth angle of −45 °, and the azimuth angle of the linear polarizer 15 ′ is set to 0 ° and 90 °. In addition, at the azimuth angle of each linear polarizer 15 ′, two types of polarized light that is symmetrical on the origin on the Poincare sphere PB are emitted. Thus, Thus, to obtain the following equation [82] - [Expression 87], and further, in this [Expression 82] - [Expression _87], by determining the _{r 30, r 31, r 32} , r ₃₃ is obtained as in the following [Equation 88] to [Equation 90].

On the other hand, r ₀₁ , r ₀₂ and r ₀₃ are calculated in common as in the following [Equation 91] to [Equation 93].

As described above, if r ₁₀ , r ₂₀ , and r ₃₀ can be determined, the Mueller matrix (R) of the entire optical system including the illumination optical system 3 and the photomask 5 is referred to as the Mueller matrix (M) of the projection optical system 6. Can be identified independently. These r ₁₀ , r ₂₀ , and r ₃₀ are expressed by the following [Equation 94] to [Equation 96] from the above relational expression.

  However, each coefficient was set as in the following [Equation 97].

When the Mueller matrix (R) of the optical system composed of the illumination optical system 3 and the photomask 5 is obtained in this way, this inverse matrix (R ⁻¹ ) is used to obtain the method shown in the first embodiment. The measured Mueller matrix (M ′) of the entire optical system from the illumination optical system 3 to the projection optical system 6 can be corrected as follows.

As mentioned above, although embodiment of this invention was described, this invention is not limited to these, A various change, addition, etc. are possible in the range which does not deviate from the meaning of invention. For example, in the above embodiment, the three polarized lights (S ⁽¹⁾ ) to (S ⁽³⁾ ) that are orthogonal to each other on the Poincare sphere PB, and the polarized light that is symmetrical with the origin on the Poincare sphere PB. Although six types of polarized light of (S ⁽⁴⁾ ) to (S ⁽⁶⁾ ) are incident, the present invention is not limited to this. For example, in addition to the polarized light (S ⁽¹⁾ ) to (S ⁽³⁾ ) having ^three orthogonal relationships, any one of the polarized lights (S ⁽⁴⁾ ) to (S ⁽⁶⁾ ) It is sufficient that only the four types of polarized light are incident on the optical system as the measurement target.

Converted polarized light (S ⁽¹⁾ ′) to (S ⁽⁶⁾ ′) obtained by converting six types of projection light (S ⁽¹⁾ ) to (S ⁽⁶⁾ ) by an optical system having a Mueller matrix (M). Is expressed as follows.

For example, from (S ⁽¹⁾ ′) and (S ⁽⁴⁾ ′), the second and second rows of the Mueller matrix (M) are obtained as follows.

When two rows of the four rows of the Mueller matrix (M) are obtained in this way, the remaining two rows are the remaining two converted polarized lights (S ⁽²⁾ ) and (S ⁽³⁾ ), or (S ^{( 5)} ) and (S ⁽⁶⁾ ).

1 shows a configuration of an exposure apparatus 100 to which a first embodiment of the present invention is applied. The structure of the 1st polarization converter 2 is shown. Polarized light (S ⁽¹⁾ ), (S ⁽²⁾ ), (S ⁽³⁾ ), (S ⁽⁴⁾ ), (S ⁽⁵⁾ ), (S ^{( 6)} ) is expressed on the Poincare sphere PB. The structure of the 2nd polarization converter 7 is shown. The structure of the 2nd polarization converter 7 is shown. An example of the measurement result of the Mueller matrix (M) is shown. The structure of the exposure apparatus 100 with which the 3rd Embodiment of this invention is applied is shown. The structure of the exposure apparatus 100 with which the 4th Embodiment of this invention is applied is shown.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 ... Exposure apparatus, 1 ... Light source device, 2 ... 1st polarization converter, 3 ... Illumination optical system, 5 ... Photomask, 6 ... Projection optical system, 7 ... 2nd polarization conversion device, 8... Photodetector, 9... 3rd polarization conversion device, 10... Imaging plane, 11. Driver, 14 ... A / D converter, 15, 15 '... Linear polarizer, 16 ... 1/4 wavelength plate, 21 ... 1/2 wavelength plate, 22 ... 1 / 4 wavelength plate, PB ... Poincare sphere, 71 ... 1/4 wavelength plate, 72 ... Linear polarizer.

Claims (12)

A light source that emits a luminous flux;
An illumination optical system for guiding the luminous flux to a photomask;
In an exposure apparatus comprising: a projection optical system for projecting an image of a mask pattern onto the object to be exposed;
A first polarization conversion device capable of sequentially converting the light flux into any one of a plurality of polarized light beams that are orthogonal to each other on a Poincare sphere;
The light intensity of the converted polarized light after being placed near the imaging position of the image of the mask pattern by the projection optical system after the polarized light passes through the illumination optical system, the photomask, and the projection optical system. A light intensity detector for detecting;
A second polarization conversion device including a linear polarizer and a linear phase shifter and inserted into the optical path of the converted polarized light;
A plurality of light obtained by acquiring a plurality of data of the light intensity detected by the light intensity detector by switching an azimuth angle of the linear polarizer and an insertion / removal of the linear phase shifter or an azimuth angle An exposure apparatus comprising: a parameter calculation unit that calculates a Stokes parameter of the converted polarized light based on intensity data.   The exposure apparatus according to claim 1, wherein the linear polarizer can be set to have an azimuth angle of 45 × n ° (n is an integer). The parameter calculator is
At the respective azimuth angles, the linear phase shifter is separated or the azimuth angles of the linear polarizer and the linear polarizer are matched, and the azimuth angles of the linear polarizer are set to 0 °, 90 °, 45 °, and −45 °. Obtaining light intensity data detected by the light intensity detector;
The linear phase shifter is inserted at a predetermined azimuth angle different from the azimuth angle of the linear polarizer, and the azimuth angle of the linear polarizer is switched between 0 ° and 90 °, or between 45 ° and −45 °. The light intensity data detected by the light intensity detector before and after this switching is acquired,
2. An exposure apparatus according to claim 1, wherein a Stokes parameter of the converted polarized light is calculated based on the data of the light intensity. The parameter calculator is
The linear phase shifter is removed or the azimuth angle of the linear polarizer is matched with the azimuth angle of the linear polarizer, and the azimuth angle of the linear polarizer is sequentially set to 0 ° and 90 °. The detected light intensity data is obtained, and based on these light intensity data, the Stokes parameters (s ₀ ′, s ₁ ′, s ₂ ′, s ₃ ′) of the converted polarized light s ₀ ′ and Calculate s ₁ '
The linear phase shifter is separated or the azimuth angle of the linear polarizer is made coincident, and the azimuth angle of the linear polarizer is sequentially set to 45 ° and −45 °, and the light intensity detection unit at each azimuth angle. in obtaining data of the detected light intensity, based on the data of these light intensities, the Stokes parameters of the converted polarized light _{(s 0 ', s 1'} , s 2 ', s 3') of s ₀ ' And s ₂ '
The linear phase shifter is inserted at a predetermined azimuth angle different from the azimuth angle of the linear polarizer, and the azimuth angle of the linear polarizer is switched between 0 ° and 90 °, or between 45 ° and −45 °. Thus, before and after this switching, data on the light intensity detected by the light intensity detector is acquired, and based on the data on the light intensity, the Stokes parameters (s ₀ ′, s ₁ ′, _2. The exposure apparatus according to claim 1, wherein s ₀ ′ and s ₃ ′ of s ₂ ′, s ₃ ′) are calculated.   The exposure apparatus according to claim 1, wherein the linear phase shifter is a ¼ wavelength plate. The quarter-wave plate has a first orientation when calculating s ₀ ′ and s ₃ ′ of Stokes parameters (s ₀ ′, s ₁ ′, s ₂ ′, s ₃ ′) of the converted polarized light. And a second azimuth angle that is 90 ° different from the first azimuth angle, and the parameter calculation unit obtains light intensity data at each azimuth angle, and a Stokes parameter of the converted polarized light. 6. An exposure apparatus according to claim 5, wherein The luminous flux emitted from the light source is linearly polarized light,
The first polarization conversion device includes a polarization rotation element and a quarter wavelength plate,
The quarter-wave plate is inserted so that the azimuth is 45 ° with respect to the polarization direction of the light beam that has passed through the polarization rotation element when converting linearly polarized light into circularly polarized light, and in other cases The exposure apparatus according to claim 1, wherein the light beam is inserted such that the direction of change of the light flux coincides with the azimuth angle, or is removed from the optical path of the light flux.   The exposure apparatus according to claim 7, wherein the polarization rotation element is a half-wave plate configured to change an azimuth angle every 22.5 degrees with respect to the light flux.   The exposure apparatus according to claim 7, wherein the polarization rotation element is a half-wave plate configured to change an azimuth angle every 45 degrees with respect to the light flux.   The exposure apparatus according to claim 7, wherein the polarization rotation element includes a depolarization element and a linear polarizer capable of changing an azimuth angle.   A third polarization conversion device including a linear polarizer and a linear phase shifter whose azimuth angle can be set to 45 × n ° (n is an integer) is disposed between the photomask and the projection optical system. The exposure apparatus according to claim 1, wherein the exposure apparatus is configured to be able to perform the operation.   The parameter calculation unit switches the azimuth angle of the linear polarizer of the third polarization conversion device and the insertion / removal of the linear phase shifter of the third polarization conversion device or the azimuth angle to a plurality of states, and In each state, a plurality of types of polarized light that is symmetrical with respect to the origin on the Poincare sphere are emitted from the first polarization conversion device to obtain data on a plurality of light intensities from the light intensity detector. 12. The exposure apparatus according to claim 11, wherein a parameter of the Mueller matrix of the photomask and the illumination optical system is calculated based on a plurality of types of light intensity data.