JP2006250859A - Surface shape measuring method, surface shape measuring instrument, projection optical system manufacturing method, projection optical system, and projection exposure device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a surface shape measuring method and a surface shape measuring instrument with the theory of PDI applied thereto for measuring the shape of a surface under inspection with high accuracy independently of surface accuracy error of a reflective surface for bending an optical path of a light beam under inspection reflected by the surface. <P>SOLUTION: According to this surface shape measuring method, a measuring system is used for: illuminating the surface 5a under inspection with a part of an ideal spherical wave projected from a point light source 3a; causing the optical path of the light beam LW under inspection reflected by the surface 5a to interfere with another part of the spherical wave by bending the optical path by the reflective surface 3; and acquiring data on interference fringes thereby generated as shape data on the surface. This measuring method is characterized in that an area of the surface illuminated by the spherical wave is limited to a partial area by a mask M to acquire shape data on the partial area. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

The present invention relates to a projection exposure apparatus used for projection exposure such as EUVL (EUVL: Extreme UltraViolet Lithography), a projection optical system mounted thereon, and a method of manufacturing the projection optical system.
The present invention also relates to a surface shape measuring method for measuring the shape of an optical surface such as a mirror or a lens of a projection optical system, and a surface shape measuring device applied thereto.

Conventionally, Fizeau interferometers and Twiman-Green interferometers have been used for measuring the shape of optical surfaces such as lenses and mirrors. Since all of these interferometers require a reference surface, the shape data acquired by the interferometer merely represents a comparison between the test surface and the reference surface. Therefore, the measurement accuracy cannot exceed the surface accuracy of the reference surface.
Therefore, PDI (Point-Diffraction-Interferometer) that does not require a reference plane has been proposed (Patent Documents 1, 2, etc.). In PDI, a point light source is generated by an optical member such as a pinhole mirror or an optical fiber, and the shape of the test surface is measured with reference to an ideal spherical wave emitted from the point light source.

For example, in the PDI described in Patent Document 2, an ideal spherical wave is generated by a pinhole mirror, and the test surface is illuminated with a part of the ideal spherical wave. The light beam reflected by the test surface (test light beam) is collected in the vicinity of the pinhole of the pinhole mirror and then reflected by the surface of the pinhole mirror to bend the optical path. The test light beam interferes with another part of the ideal spherical wave to form interference fringes. The interference fringe data is acquired by the CCD image sensor as shape data. As described above, the shape data acquired by PDI represents a comparison between the test surface and the wavefront of an ideal spherical wave, and thus is highly accurate.
JP-A-2-228505 JP-A-6-17447

However, when the surface accuracy error of the test surface is large, or when the design shape of the test surface is an aspherical surface, high-precision measurement is difficult even with this PDI. In particular, it is difficult to measure the fine shape component of the test surface.
This is because if the aspherical amount of the test surface is large, the test light beam reflected by the test surface is not completely condensed in the vicinity of the pinhole of the pinhole mirror but forms a larger spot. At this time, the test light beam is diffracted and reflected according to the surface accuracy error of the pinhole mirror. For this reason, not only the necessary light beam (= the test light beam specularly reflected by the pinhole mirror and the ideal spherical wave) but also the noise light beam (= the test light beam diffracted and reflected by the pinhole mirror) is incident on the CCD. Resulting in.

Therefore, the present invention is based on the principle of PDI that can measure the shape of the test surface with high accuracy without depending on the surface accuracy error of the reflection surface for bending the optical path of the test light beam reflected by the test surface. It is an object of the present invention to provide a surface shape measuring method and a surface shape measuring device that are used.
Another object of the present invention is to provide a projection optical system manufacturing method, a high-performance projection optical system, and a high-performance projection exposure apparatus that can manufacture a high-performance projection optical system.

  The surface shape measuring method of the present invention illuminates a test surface with a part of an ideal spherical wave emitted from a point light source, and bends the optical path of a test light beam reflected by the test surface by the reflection surface. In a surface shape measurement method using a measurement system that causes interference fringe data to be caused to interfere with another part of a target spherical wave and obtains data of interference fringes as shape data of the target surface, the test by the ideal spherical wave The illumination area of the surface is limited to a partial area, and shape data of the partial area is acquired.

In the surface shape measuring method of the present invention, the measurement system may be provided with a pinhole mirror serving as both the means for generating the point light source and the reflecting surface.
Further, in the surface shape measuring method of the present invention, the position of the partial area is moved on the test surface, the shape data of a plurality of different partial areas on the test surface are acquired, and they are connected. Then, shape data of the entire area of the test surface may be created.

Further, in the surface shape measuring method of the present invention, in the connection, the environmental change of the measurement system (which is a factor of an error that varies with time, for example, drift of the measurement system (light quantity fluctuation, etc.), measurement system, etc. The error generated in the shape data of the plurality of partial areas due to the change in the posture of the surface to be measured) may be corrected.
Further, in the surface shape measurement method of the present invention, the shape data of the entire area of the test surface obtained by collectively illuminating the entire area of the test surface is compared with the shape data of the plurality of partial areas. The error may be obtained.

In the surface shape measurement method of the present invention, the difference between the low frequency shape component included in the shape data of the entire surface of the test surface and the low frequency shape component included in the shape data of the plurality of partial regions. The error may be obtained based on the above.
Moreover, in the surface shape measuring method of this invention, the design shape of the said to-be-tested surface may be an aspherical surface.

  Further, the surface shape measuring apparatus of the present invention illuminates the test surface with a part of the ideal spherical wave emitted from the point light source and bends the optical path of the test light beam reflected by the test surface at the reflection surface. In a surface shape measuring apparatus having a measurement system that interferes with another part of the ideal spherical wave and obtains data of interference fringes generated thereby as shape data of the test surface, the incident side of the test surface And a mask for limiting the illumination area of the test surface by the ideal spherical wave to a partial area.

In the surface shape measuring apparatus of the present invention, the measurement system may be provided with a pinhole mirror that serves as the point light source and the reflecting surface.
Moreover, in the surface shape measuring apparatus of this invention, you may provide the mechanism which moves the position of the said partial area | region on the said to-be-tested surface.
Further, the surface shape measuring apparatus of the present invention may include a calculation unit that connects shape data of a plurality of different partial areas on the test surface to create shape data of the entire area of the test surface. .

Further, in the surface shape measuring apparatus of the present invention, the calculation means may correct an error generated in the shape data of the plurality of partial regions due to an environmental change of the measurement system in the connection. .
Further, in the surface shape measuring apparatus of the present invention, the calculation means includes shape data of the entire area of the test surface acquired by collectively illuminating the entire area of the test surface, and shape data of the plurality of partial areas. And the error may be obtained.

Further, in the surface shape measuring apparatus of the present invention, the calculation means includes a low frequency shape component included in the shape data of the entire surface of the test surface and a low frequency included in the shape data of the plurality of partial regions. The error may be obtained based on a difference from the shape component.
Further, in the surface shape measuring apparatus of the present invention, among the surface accuracy errors of the reflecting surface, at least the pitch of the waviness component to be excluded from the influence on the shape data is P, and the curvature of the approximate spherical surface of the test surface is When the radius is R and the light source wavelength of the measurement system is λ, the size d of the opening of the mask may satisfy the formula d <R · λ / P.

Moreover, in the surface shape measuring apparatus of this invention, the mask which can change the position of an opening part may be sufficient as the said mask.
Moreover, in the surface shape measuring apparatus of this invention, the mask which can change the size and / or shape of an opening part may be sufficient as the said mask.
According to the projection optical system manufacturing method of the present invention, according to the procedure of measuring at least one optical surface of the projection optical system by the surface shape measuring method according to any one of the present invention, and the result of the measurement And a procedure for processing the optical surface.

The projection optical system of the present invention is manufactured by the method for manufacturing a projection optical system of the present invention.
The projection exposure apparatus of the present invention is equipped with the projection optical system of the present invention.
In the projection exposure apparatus of the present invention, the exposure light may be EUV light having a wavelength of 50 nm or less.

According to the present invention, the principle of PDI capable of measuring the shape of the test surface with high accuracy without depending on the surface accuracy error of the reflection surface for bending the optical path of the test light beam reflected by the test surface. A surface shape measuring method and a surface shape measuring apparatus using the above are realized.
In addition, according to the present invention, a manufacturing method of a projection optical system, a high-performance projection optical system, and a high-performance projection exposure apparatus capable of manufacturing a high-performance projection optical system are realized.

[First Embodiment]
A first embodiment of the present invention will be described with reference to FIGS. 1, 2, 3, 4, 5, 6, and 7.
The present embodiment is an embodiment of a surface shape measuring apparatus and a surface shape measuring method using the same.
FIG. 1 is a configuration diagram of the measurement apparatus. As shown in FIG. 1, the measurement apparatus includes a laser light source 1, a lens 2, a pinhole mirror 3, a lens 6, a CCD image sensor 7, a mask M, moving mechanisms 8 and 9, a control circuit 21, a computer 22, and the like. Provided. In FIG. 1, reference numeral 5 indicates a test object 5 such as an aspherical mirror. The test surface 5a of the test object 5 is hereinafter referred to as “test aspheric surface 5a” because the design shape is aspheric.

As shown in FIG. 1, in this measuring apparatus, a pinhole mirror 3 is disposed between the test object 5 and the laser light source 1 (details of the pinhole mirror 3 will be described later).
The light beam emitted from the laser light source 1 is collected by the lens 2 and irradiates a pinhole 3 c provided in the pinhole mirror 3. Due to the diffractive action of the pinhole 3c, an ideal spherical wave SW is generated on the exit side of the pinhole mirror 3.

  A part of the ideal spherical wave SW illuminates the test aspheric surface 5a. The reflected light generated on the test aspheric surface 5a is a test light beam LW having a wavefront WW deformed according to the shape of the test aspheric surface 5a. The test light beam LW travels toward the pinhole mirror 3, is reflected near the pinhole 3 c, bends its optical path, and travels toward the lens 6. The test light beam LW incident on the lens 6 reaches the image pickup surface 7a of the CCD image pickup device 7 as a substantially parallel light beam.

The lens 6 plays a role of forming an image of the test aspheric surface 5a on the imaging surface 7a, and its distortion is sufficiently reduced. Therefore, it can be considered that each position on the imaging surface 7a has a one-to-one correspondence with each position on the test aspheric surface 5a.
The other part of the ideal spherical wave SW goes to the lens 6 together with the test light beam LW as the reference light beam LR, and reaches the imaging surface 7a of the CCD image pickup device 7 as a parallel light beam.

Therefore, interference fringes are generated on the imaging surface 7a due to the test light beam LW and the reference light beam LR. The phase distribution of the interference fringes represents the difference between the wavefront WW of the test light beam LW and the wavefront (ideal spherical surface) of the reference light beam LR, that is, the shape of the test aspheric surface 5a.
The CCD image sensor 7 captures the interference fringes and acquires luminance distribution data of the interference fringes. This luminance distribution data is sent to the computer 22 via the control circuit 21. The computer 22 analyzes the luminance distribution data of the interference fringes by a known analysis method, and converts it into shape data of the test aspheric surface 5a. The computer 22 is preinstalled with a program for this analysis.

  In addition, a well-known phase shift interferometry can be applied to the above measurement apparatus. In the phase shift interferometry, luminance distribution data is repeatedly acquired while moving the test object 5 minutely in the optical axis direction by a moving mechanism 9 such as a piezo element. The movement amount of the test object 5 by the moving mechanism 9 at this time and the imaging timing of the interference fringes by the CCD image sensor 7 are controlled by the control circuit 21. The computer 22 acquires the shape data of the test aspheric surface 5a by substituting the brightness distribution data group acquired in this way into a predetermined arithmetic expression. The shape data acquired by such a phase shift method represents the shape of the aspheric surface 5a to be detected with high accuracy (the phase shift method is well known, and the details are omitted).

  In this measuring apparatus, the mask M is supported by a moving mechanism 8 such as a stage, and can be inserted into and removed from the optical path immediately before the aspheric surface 5a to be examined. The moving mechanism 8 can also change the position of the opening H of the mask M on the test aspheric surface 5a by sliding the mask M in a direction perpendicular to the optical axis. The sliding amount of the mask M by the moving mechanism 8 is controlled by the control circuit 21 (details of the mask M will be described later).

FIG. 2 is an enlarged view of the periphery of the pinhole mirror 3 when the mask M is separated from the optical path.
As shown in FIG. 2, the pinhole mirror 3 is formed by forming a metal film 3b made of chromium or the like on the surface of a parallel flat glass substrate 3a. An opening to be a pinhole 3c is provided in the approximate center of the metal film 3b by etching or the like.

The pinhole mirror 3 has the glass substrate 3a facing the laser light source 1 side.
The test light beam LW from the test aspheric surface 5a forms a relatively large spot in the vicinity of the pinhole 3c in the metal film 3b. The size of the focused spot is increased because the shape of the wavefront WW of the test light beam LW is aspherical.
At this time, the test light beam LW is diffracted and reflected according to the surface accuracy error of the metal film 3b. That is, the test light beam LW reflected by the metal film 3b and having its optical path bent has a zero-order diffraction component (a component composed of specularly reflected zero-order diffracted light) LW ₀ and a + 1st-order diffraction component (diffracted and reflected + 1st-order light). Component) LW ₊₁ and −1st order diffraction component (component consisting of diffracted and reflected −1st order diffracted light) LW ₋₁ .

FIG. 3 is a diagram for explaining a state in the vicinity of the imaging surface 7a at this time.
As shown in FIG. 3, the 0th-order diffraction component LW ₀ , the + 1st-order diffraction component LW ₊₁ , and the −1st-order diffraction component LW ₋₁ emitted from the pinhole mirror 3 are incident on areas shifted from each other on the imaging surface 7a. To do.
In the imaging surface 7a, the + 1-order diffraction components LW ₊₁ wavefront WW ₊₁ and -1 order diffraction component LW _-1 wavefront WW _-1, and lateral with respect to the wavefront WW ₀ of zero-order diffraction components LW ₀ superimposed To do. Therefore, an error occurs in the luminance distribution data including shape information of the aspheric surface 5a (luminance distribution data of the interference fringes formed by the wavefront WW _0).

FIG. 4 is a diagram illustrating the mask M.
As shown in FIG. 4A, a square-shaped opening H is provided in the approximate center of the mask M, and the other region is a light shielding portion.
When this mask M is inserted into the optical path, as shown in FIG. 4B, the illumination area by this measuring device is limited to only a square partial area E _i smaller than the entire measurement area E of the aspheric surface 5a to be examined. be able to.

Here, as shown in FIG. 4B, the size of the opening H of the mask M is set to individual partial areas E _i (i = 1, 1) obtained by dividing the square in which the entire measurement area E is inscribed into nine equal parts. 2,..., 9).
If this mask M is moved and the position of the opening H on the aspheric surface 5a to be examined is slid, the illumination area by this measuring apparatus is changed to nine partial areas E ₁ , E ₂ , E ₃ , E ₄ , _{E 5, E 6, E 7} , can switch between E _8, E _9.

FIG. 5 is a diagram illustrating a state in the vicinity of the imaging surface 7a when the mask M is inserted in the optical path.
As shown in FIG. 5, even if the mask M is inserted in the optical path, the 0th-order diffraction component LW ₀ , the + 1st-order diffraction component LW ₊₁ , and the −1st-order diffraction component LW ₋₁ of the test light beam LW The light is incident on the regions 7a shifted from each other. However, since the illumination area on the aspheric surface 5a to be examined is limited to a partial area, the sizes of the wavefronts WW ₊₁ , WW ₋₁ , and WW ₀ are small.

Thus, the illumination area far been limited sufficiently small, the wavefront WW ₊₁ on the imaging surface 7a, a wave front WW _-1, and the wave front WW ₀ completely separated. If separated in this way, no error occurs in the luminance distribution data including the shape information of the partial region (the luminance distribution data of the interference fringes formed by the wavefront WW ₀ ).
For this reason, the size of the opening H of the mask M of the present measurement apparatus is such that the formation area of the wave fronts WW ₊₁ and WW _{-1 and} the formation area of the wave front WW ₀ are completely separated on the imaging surface 7a. It is set to a sufficiently small size.

However, the deviation amounts of the wave fronts WW ₀ , WW ₊₁ , WW ₋₁ depend on the pitch P of the undulation component of the surface accuracy error of the pinhole mirror 3 (surface accuracy error of the metal film 3b), and the undulation component is fine. The larger it is (that is, the smaller the pitch P of the swell component), the larger. On the other hand, it is considered that the surface accuracy error of the pinhole mirror 3 includes various undulation components having different pitches.
For this reason, when the manufacturer of the measurement apparatus sets the size of the opening H of the mask M, it is necessary to consider the pitch P of the roughest undulation component among the undulation components whose influence on the measurement result should be removed. There is. That is, the size of the opening H of the mask M needs to be set so that the wavefronts WW ₀ , WW ₊₁ , and WW ₋₁ generated due to the roughest swell component are completely separated on the imaging surface 7a. There is. This will be specifically described below.

FIG. 6 is a diagram for explaining the influence of the undulation component of the pitch P generated in the pinhole mirror 3.
First, as shown by a solid line in FIG. 6, attention is paid to a certain test light emitted from a coordinate y (coordinate based on the optical axis OA) on the test aspheric surface 5a. In FIG. 6, the dotted line indicates one of the ± first-order diffraction components, and the alternate long and short dash line indicates the optical axis of each optical element.

The angle θ formed by the test light with the optical axis OA of the test aspheric surface 5a satisfies the following expression (1).
y = R · sin θ (1)
Here, R is the radius of curvature of the approximate spherical surface of the aspheric surface 5a to be examined.
Further, the angle θ ′ formed by the diffracted reflection of the test light by the pinhole mirror 3 and the ± 1st-order diffraction component generated thereby with the optical axis OA of the lens 6 is a swell component of the surface accuracy error of the pinhole mirror 3 The following formula (2) is satisfied together with the pitch P.

sin (φ + θ) ± λ / P = sin (φ + θ ′) (2)
Here, φ is a common angle formed by the optical axis OA of the aspheric surface 5a to be tested and the lens 6 and the normal line of the pinhole mirror 3, and λ is a light source wavelength of the measuring apparatus.
Further, the coordinate y ′ (coordinate based on the optical axis OA) on the imaging surface 7a on which the ± first-order diffraction component traveling at the angle θ ′ enters after passing through the lens 6 satisfies the following expression (3).

y ′ = f · sin θ ′ (3)
Here, f is the focal length of the lens 6.
Therefore, in order to eliminate the influence of the waviness component finer than the pitch P, the angles θa and θb of the light rays incident on the edge of the opening H of the mask M satisfy the following conditional expression (4). Good.

sin (φ + θa) + λ / P> sin (φ + θb) (4)
Here, considering an actual measuring apparatus, the spread of the ideal spherical wave SW emitted from the pinhole mirror 3 is not so large, so the angle φ is set to a small value. If the NA of the test aspheric surface 5a is 0.2, φ is set to about 15 deg. Therefore, it is assumed here that φ is small, and conditional expression (4) is approximated as conditional expression (5).

θb−θa <λ / P (5)
On the other hand, the size d of the opening H of the mask M is expressed by Expression (6) by the angles θa and θb of the light rays incident on the edge of the opening H and the curvature radius R of the approximate spherical surface of the aspheric surface 5a to be tested. The
d = R · (θb−θa) (6)
Therefore, from the expressions (5) and (6), the size d of the opening H of the mask M that eliminates the influence of the swell component finer than the pitch P may satisfy the following conditional expression (7).

d <R · λ / P (7)
For example, in order to eliminate the influence of the waviness component finer than the pitch P = 8 μm when the light source wavelength λ = 633 nm and the radius of curvature R of the approximate spherical surface of the aspheric surface 5a to be tested are 400 mm, the opening of the mask M The size d of H should just satisfy | fill the formula of d <400 * 0.633 / 8 = 32mm.

FIG. 7 is a flowchart showing the procedure of the surface shape measuring method using the above measuring apparatus.
In step S1, the mask M is removed from the optical path, and the illumination area by this measurement apparatus is set to the entire measurement area E of the aspheric surface 5a to be examined. In this state, the measurement apparatus acquires the luminance distribution data T ₀ for all the measurement areas E at once. The computer 22 converts the luminance distribution data T ₀ of the entire measurement region E into shape data A ₀ . Note that phase shift interferometry may be applied to step S1.

In step S2, the computer 22 expands the shape data A ₀ into a Zernike polynomial, and extracts a low-order term coefficient value (low frequency shape component) B ₀ . This shape component B ₀ represents the rough shape of the entire measurement region E.
In step S3, inserting the mask M in the optical path, the partial regions E ₁ of the aspheric surface 5a of the area illuminated by the measuring device, E _2, E _3, · · ·, sequentially sets the E _9. This measuring device includes _first illumination area each partial region _{E, E 2, E 3,} ···, each state set in E _9, partial regions _{E 1, E 2, E 3} , ···, E ₉ luminance distribution data T ₁ , T ₂ , T ₃ ,..., T ₉ are acquired individually.

Incidentally, the luminance distribution data _{T 1, T 2, T 3} , ···, the T ₉ each subregion _{E 1, E 2, E 3} , ···, the luminance distribution of a slightly larger area than the respective E ₉ It is data. Therefore, the luminance distribution data T ₁ , T ₂ , T ₃ ,..., T ₉ slightly overlap each other.
Then, the computer 22, these luminance distribution data _{T 1, T 2, T 3} , converts ..., the shape data A ₁ to T _9, respectively, A _2, A _3, ..., a A _9. Note that phase shift interferometry may be applied to step S3.

In step S4, the computer 22 expands the shape data A ₁ , A ₂ , A ₃ ,..., A ₉ into Zernike polynomials, respectively, and low-order term coefficient values (low frequency shape components) B ₁ , B _2. , B ₃ ,..., B ₉ are extracted. These shapes components _{B 1, B 2, B 3} , ···, B 9 is the partial regions _{E 1, E 2, E 3} , ···, with represents the general shape of E _9, the It includes an error (consisting of a posture error, a tilt component, and a shift component) due to the posture fluctuation of the inspection object 5. This posture variation is one type of “environmental change” in the claims.

In step S5, the computer 22, the partial regions _{E 1, E 2, E 3} , ···, shape component B ₁ representing the general shape of _{E 9, B 2, B 3} , ···, the B ₉ Compared with each part of the shape component B ₀ representing the rough shape of the entire measurement region E, the tilt component and the shift component of the posture error are obtained from the difference between the two. The computer 22 corrects the shape data A ₁ , A ₂ , A ₃ ,..., A ₉ so that the tilt component and shift component are eliminated, and the shape data C ₁ , C ₂ , C ₃ ,. - to obtain a C _9.

In step S6, the computer 22 joins the corrected shape data C ₁ , c ₂ , C ₃ ,..., C ₉ to create the shape data D of the entire measurement region E of the aspheric surface 5a to be examined.
However, the shape data C ₁ , c ₂ , C ₃ ,..., C ₉ include errors due to fluctuations in the amount of light of this measuring apparatus. Therefore, when the shape data C ₁ , c ₂ , C ₃ ,..., C ₉ are simply connected, a step is generated at the connection part. This light quantity variation is one type of “environmental change” in the claims.

Therefore, the computer 22 in step S6 performs weighted averaging of the overlapping data among the shape data C ₁ , C ₂ , C ₃ ,..., C ₉ in order to smooth the joined portion.
Code S61 in FIG. 7 represents the step of obtaining shape data C ₁₂ by connecting to the weighted average of the shape data C _1, C _2.

The value of the weight function W ₁ to be multiplied by the shape data C ₁ by the weighted average in step S61 is 1 in the region on the shape data C ₁ side that is out of the overlap region, as shown by the solid line in FIG. In the overlap region, a continuous value between ₁ and 0 is taken from the shape data C ₁ side to the shape data C ₂ side, and 0 in the region on the shape data C ₂ side that is out of the overlap region.

Further, the value of the weight function W ₂ to be multiplied by the shape data C ₂ by the weighted average in step S61 is 1 in the region on the shape data C ₂ side outside the overlap region, as indicated by the dotted line in FIG. In the overlap region, the continuous value is ₁ to 0 from the shape data C ₂ side to the shape data C ₁ side, and 0 in the region on the shape data C ₁ side that is out of the overlap region.

By using such weight functions W ₁ and W ₂ , the shape data C ₁ and C ₂ can be smoothly joined without a step.
Code S62 in FIG. 7 represents the step of obtaining shape data C ₁₂₃ by connecting to the weighted average of shape data C _12, C _3.
The value of the weight function W ₁₂ to be multiplied by the shape data C ₁₂ by the weighted average in step S62 is 1 in the region on the shape data C ₁₂ side that is out of the overlap region, as shown by the solid line in FIG. In the overlap region, a continuous value between 1 and 0 is taken from the shape data C ₁₂ side to the shape data C ₂ side, and 0 in the region on the shape data C ₃ side that is out of the overlap region.

Further, the value of the weight function W ₃ to be multiplied by the shape data C ₃ by the weighted average in step S62 is 1 in the region on the shape data C ₃ side that is out of the overlap region, as indicated by a dotted line in FIG. , and the in the overlapping region becomes a continuous value between 1-0 from the shape data C ₃ side to the shape data C ₁₂ side, it becomes 0 in the area of out of the overlapping area shape data C ₁₂ side.

By using such weight functions W ₁₂ and W ₃ , the shape data C ₁₂ and C ₃ can be smoothly joined without any step.
Similarly, the computer 22 connects the other shape data C ₄ , C ₅ ,..., C ₉ to complete the shape data D of the entire measurement region E of the test aspheric surface 5a (step S7).
In step S6, it is desirable to perform joining in a predetermined direction (X direction) and then perform joining in a direction (Y direction) orthogonal thereto. For example, the order is (1), (2), (3), (4).

(1) Shape data C ₁₂₃ is obtained by connecting the shape data C ₁ , C ₂ , and C ₃ .
(2) Shape data C ₄₅₆ is obtained by connecting the shape data C ₄ , C ₅ , and C ₆ .
(3) Shape data C ₇₈₉ is obtained by connecting the shape data C ₇ , C ₈ , and C ₉ .
(4) Shape data D is obtained by connecting shape data C ₁₂₃ , C ₄₅₆ , and C ₇₈₉ .
As described above, according to this measuring device, the partial area E ₁ of the illumination area of the aspheric surface 5a with a mask _{M, E 2, E 3,} ···, limited to E _9, they subregion E _1, E _2, E _3, ···, the luminance distribution data T ₁ for each _{E 9, T 2, T 3} , ···, it is possible to obtain a T _9. Each of the luminance distribution data T ₁ , T ₂ , T ₃ ,..., T ₉ acquired individually is not affected by the surface accuracy error of the pinhole mirror 3.

Further, in this measurement method, the partial regions _{E 1, E 2, E 3} , ···, the shape data A ₁ of _{E 9, T 2, T 3} , ···, Upon joining the T _9, the test The error due to the posture change of the object 5 is corrected (steps S1 to S5 in FIG. 7), and the influence of the error due to the light amount fluctuation of the measuring apparatus is suppressed (step S6 in FIG. 7). Therefore, the shape data D of the entire measurement region E of the test aspheric surface 5a can be obtained with high accuracy.

In this measuring apparatus, among the test light beam LW reflected by the test aspheric surface 5a, in particular, the light beam generated due to the fine shape component (high frequency shape component) of the test aspheric surface 5a is: Compared with other light beams, the pinhole mirror 3 has a low degree of light condensing (the light condensing spot becomes large), so that it is greatly affected by the surface accuracy error of the pinhole mirror 2. For this reason, conventionally, it has been difficult to measure particularly fine shape components of the aspheric surface 5a to be examined with high accuracy. However, in the measuring apparatus and the measuring apparatus, the surface accuracy influence the luminance distribution data T ₁ of the error of the pinhole mirror _{2, T 2, T 3,} ···, because it is excluded from the T _9, fine shape Even components can be measured with high accuracy.

  Further, in this measurement apparatus, the mask M whose opening size and opening position are fixed is used and the entire mask M is moved. However, even if a mask whose opening size and opening position is variable is used. Good. For example, as shown in FIG. 8, in the mask formed by combining a pair of L-shaped blinds B1 and B2, the size and position of the opening H can be freely changed simply by driving the blinds B1 and B2. In addition, as such a mask, what was disclosed by Unexamined-Japanese-Patent No. 6-324474 etc. is applicable.

In that case, a measurement method in which the opening size is changed according to the opening position can also be applied.
Further, in this measuring apparatus, the shape of the opening H of the mask M is a square shape, but may be another shape such as a rectangle or a circle.
Further, in this measuring apparatus, the surface of the metal film 3b of the pinhole mirror 3 is used as a reflection surface for bending the optical path of the test light beam LW, but the reflectance of the test light beam LW on the surface of the metal film 3b. A reflective film may be formed to improve the above.

  Moreover, in this measuring apparatus, it may replace with the pinhole mirror 3, and may use another optical member which performs the same effect | action. For example, Japanese Patent Laid-Open No. 6-173337 discloses an interference measuring apparatus using an optical fiber or an optical waveguide instead of the pinhole mirror 3. Even in such a case, according to the present invention, the shape of the test surface can be measured with high accuracy without depending on the surface accuracy error of the exit end face of the optical fiber and the surface accuracy error of the exit end face of the optical waveguide.

[Second Embodiment]
The second embodiment of the present invention will be described below with reference to FIGS.
The present embodiment is an embodiment of a method for manufacturing a projection optical system. The projection optical system manufactured in this embodiment is, for example, a projection optical system PL mounted on a projection exposure apparatus for EUVL as shown in FIG.

  As shown in FIG. 9, an illumination optical system 101, a reflective reticle R, a projection optical system PL, and a wafer W are arranged in a projection exposure apparatus for EUVL. Reticle R is supported by reticle stage 102, and wafer W is supported by wafer stage 106. Reticle stage 102 and wafer stage 106 are driven by drive circuits 102c and 106c. The drive circuits 102c and 106c are controlled by the control unit 109.

The light source of the illumination optical system 101 emits EUV light (extreme ultraviolet light) having a wavelength of 50 nm or less, for example, 13.5 nm EUV light. The projection optical system PL is a reflection type projection optical system in which a plurality of mirrors PL1, PL2, PL3, PL4, PL5, and PL6 that can reflect EUV light emitted from a light source are sequentially arranged.
It should be noted that the optical surface inside the illumination optical system 101 and the reticle R are also given a characteristic capable of reflecting EUV light emitted from the light source.

FIG. 10 is a flowchart showing a procedure of a method for manufacturing the projection optical system PL.
In step S101, the optical design of the projection optical system PL is performed. In step S101, the surface shapes of the mirror mirrors PL1, PL2, PL3, PL4, PL5, and PL6 in the projection optical system PL are determined.
In the next step S102, each mirror PL1, PL2, PL3, PL4, PL5, PL6 is processed.

In the next step S103, the surface shape of each processed mirror PL1, PL2, PL3, PL4, PL5, PL6 is measured.
In the next step S104, it is determined whether or not the surface accuracy errors of the mirrors PL1, PL2, PL3, PL4, PL5, and PL6 are within an allowable range. If not, the process returns to step S102 and rework is performed. Is done.

The above steps S102, S103, and S104 are repeated until the surface accuracy errors of the mirrors PL1, PL2, PL3, PL4, PL5, and PL6 fall within an allowable range.
Here, if it is determined in step S104 that the surface accuracy error of each of the mirrors PL1, PL2, PL3, PL4, PL5, and PL6 has decreased to some extent, in the subsequent step S103, a minute shape component that can cause flare. Measure up to. The measurement device and measurement method of the first embodiment are applied to this measurement.

In subsequent step S104, based on the measurement result, it is determined whether or not a minute shape component that causes flare remains in each of the mirrors PL1, PL2, PL3, PL4, PL5, and PL6.
Of the mirrors PL1, PL2, PL3, PL4, PL5, and PL6, those in which fine shape components remain are subjected to high-precision processing for eliminating the swell components in step S102. .

Thereafter, when the cause of flare is removed from all the mirrors PL1, PL2, PL3, PL4, PL5, and PL6, it is determined in step S104 that the surface accuracy error is within the allowable range, and the process proceeds to step S105. PL2, PL3, PL4, PL5, and PL6 are completed and the projection optical system PL is assembled.
Thereafter, while measuring the wavefront aberration of the projection optical system PL (step S106), the distance between the mirrors PL1, PL2, PL3, PL4, PL5, and PL6 is adjusted and the eccentricity is adjusted (step S108). The projection optical system PL is completed when it falls within (step S107 OK).

  As described above, in this manufacturing method, the first embodiment is applied to the surface shape measurement of the mirrors PL1, PL2, PL3, PL4, PL5, and PL6. According to the first embodiment, since the fine shape component can be measured with high accuracy, the mirrors PL1, PL2, PL3, PL4, PL5, and PL6 with high surface accuracy in which the cause of flare is surely removed are provided. Can be manufactured. Therefore, the projection optical system PL formed by assembling these mirrors PL1, PL2, PL3, PL4, PL5, and PL6 is a high-performance projection optical system PL that does not cause flare.

A projection exposure apparatus (see FIG. 9) equipped with the projection optical system PL is a high-performance projection exposure apparatus that can transfer the pattern of the reticle R onto the wafer W with high accuracy. Therefore, according to the projection exposure apparatus, a high-performance device can be manufactured.
In this manufacturing method, the measurement apparatus and the measurement method of the first embodiment are applied to the manufacture of the reflective projection optical system PL for EUVL, but the refractive projection optical system and the catadioptric projection are used. This measuring apparatus and measuring method can also be applied to the manufacture of optical systems, projection optical systems other than those for EUVL, and imaging optical systems of optical equipment other than projection exposure apparatuses. Incidentally, 1st Embodiment is applicable not only to the surface shape measurement of a mirror but to the surface shape measurement of the surface of a lens.

Examples of the first embodiment are shown below.
The wavelength λ = 0.633 μm of the laser light source 1
-Size of illumination area (D / R) = 0.2 when measuring all measurement areas E at once
· The aspheric surface 5a of the division number (the number i of partial regions _{E i) = 3 × 3 =} 9,
The size of the illumination area (D / R) = 0.08 when measuring the partial area E _i
Where D is the radial length on the test aspheric surface 5a, and R is the radius of curvature of the test aspheric surface 5a.

  According to the present embodiment described above, of the surface accuracy errors of the pinhole mirror 3, errors due to waviness components finer than the pitch P = 8 μm are eliminated.

It is a block diagram of the measuring apparatus of 1st Embodiment. It is an enlarged view of the periphery of the pinhole mirror 3 when the mask M is separated from the optical path. It is a figure explaining the mode of the imaging surface 7a vicinity when the mask M has left | separated from the optical path. It is a figure explaining the mask M. FIG. It is a figure explaining the mode of the imaging surface 7a vicinity when the mask M is inserted in the optical path. It is a figure explaining the influence of the wave | undulation component of the pitch P which has arisen in the pinhole mirror. It is a flowchart which shows the procedure of the surface shape measuring method of 1st Embodiment. It is a figure which shows the modification of the mask M. It is a block diagram of the projection exposure apparatus for EUVL. It is a flowchart which shows the procedure of the manufacturing method of projection optical system PL of 2nd Embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Laser light source 2, 6 Lens 3 Pinhole mirror 7 CCD image pick-up element M Mask 8, 9 Moving mechanism 21 Control circuit 22 Computer 5 Test object 5 Test surface (test aspherical surface)
3a Glass substrate 3b Metal film 3c Pinhole LW Test beam LR Reference beam SW Ideal spherical wave

Claims (21)

Illuminate the test surface with a part of the ideal spherical wave emitted from the point light source, and bend the optical path of the test light beam reflected by the test surface with the reflection surface to make another part of the ideal spherical wave In the surface shape measurement method using the measurement system for causing interference and obtaining data of interference fringes generated thereby as the shape data of the test surface,
A surface shape measurement method, wherein an illumination area of the surface to be examined by the ideal spherical wave is limited to a partial area, and shape data of the partial area is acquired. In the surface shape measuring method according to claim 1,
The measurement system includes
A surface shape measuring method, comprising: a pinhole mirror serving as both the means for generating the point light source and the reflecting surface. In the surface shape measuring method according to claim 1 or 2,
The position of the partial area is moved on the test surface, the shape data of a plurality of different partial areas on the test surface is acquired, and the shape data of the entire area of the test surface is created by connecting them A surface shape measuring method characterized by: In the surface shape measuring method according to claim 3,
In the connection,
A surface shape measuring method, wherein an error generated in the shape data of the plurality of partial regions due to an environmental change of the measurement system is corrected. In the surface shape measuring method according to claim 4,
A surface shape measurement characterized in that the error is obtained by comparing shape data of the entire area of the test surface acquired by collectively illuminating the entire surface of the test surface and shape data of the plurality of partial areas. Method. In the surface shape measuring method according to claim 5,
The error is obtained based on a difference between a low-frequency shape component included in the shape data of the entire surface of the test surface and a low-frequency shape component included in the shape data of the plurality of partial regions. Surface shape measurement method. In the surface shape measuring method according to any one of claims 1 to 6,
The surface shape measuring method, wherein the design shape of the test surface is an aspherical surface. Illuminate the test surface with a part of the ideal spherical wave emitted from the point light source, and bend the optical path of the test light beam reflected by the test surface with the reflection surface to make another part of the ideal spherical wave In the surface shape measuring apparatus provided with a measurement system for causing interference and obtaining data of interference fringes generated thereby as shape data of the test surface,
A surface shape measuring apparatus comprising: a mask that can be arranged on an incident side of the test surface and restricts an illumination area of the test surface by the ideal spherical wave to a partial area. In the surface shape measuring apparatus according to claim 8,
The measurement system includes
A surface shape measuring apparatus, comprising: a pinhole mirror serving as the means for generating the point light source and the reflecting surface. In the surface shape measuring apparatus according to claim 8 or 9,
A surface shape measuring apparatus comprising a mechanism for moving the position of the partial region on the surface to be examined. In the surface shape measuring apparatus according to claim 10,
A surface shape measuring apparatus comprising a calculation means for connecting shape data of a plurality of different partial areas on the test surface to create shape data of the entire area of the test surface. In the surface shape measuring apparatus according to claim 11,
The computing means is
In the joining, a surface shape measuring apparatus that corrects an error generated in the shape data of the plurality of partial regions due to an environmental change of the measurement system. In the surface shape measuring apparatus according to claim 12,
The computing means is
A surface shape measurement characterized in that the error is obtained by comparing shape data of the entire area of the test surface acquired by collectively illuminating the entire surface of the test surface and shape data of the plurality of partial areas. apparatus. In the surface shape measuring apparatus according to claim 13,
The computing means is
The error is obtained based on a difference between a low-frequency shape component included in the shape data of the entire surface of the test surface and a low-frequency shape component included in the shape data of the plurality of partial regions. Surface shape measuring device. In the surface shape measuring device according to any one of claims 8 to 14,
Of the surface accuracy error of the reflecting surface, at least the pitch of the waviness component that should be excluded from the influence on the shape data is P, the radius of curvature of the approximate spherical surface of the test surface is R, and the light source wavelength of the measurement system is When λ, the size d of the opening of the mask is
d <R · λ / P
The surface shape measuring device characterized by satisfying the formula: In the surface shape measuring apparatus according to any one of claims 8 to 15,
The said mask is a mask which can change the position of an opening part. The surface shape measuring apparatus characterized by the above-mentioned. In the surface shape measuring device according to any one of claims 8 to 16,
The said mask is a mask which can change the size and / or shape of an opening part. The surface shape measuring apparatus characterized by the above-mentioned. A procedure for measuring at least one optical surface of the projection optical system by the surface shape measuring method according to any one of claims 1 to 7,
And a step of processing the optical surface according to the result of the measurement. A projection optical system manufactured by the method for manufacturing a projection optical system according to claim 18. A projection exposure apparatus comprising the projection optical system according to claim 19. The projection exposure apparatus according to claim 20,
The exposure light is EUV light having a wavelength of 50 nm or less.

Images