JP2012181220A - MIRROR FOR ArF LITHOGRAPHY AND OPTICAL MEMBER FOR ArF LITHOGRAPHY - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an optical member having properties suitable for a mirror for an ArF lithography, and to provide a mirror for an ArF lithography using the optical member.SOLUTION: An optical member for an ArF lithography has a coefficient of thermal expansion (CTE) at 22°C of 0±200 ppb/°C, and an optical surface thereof satisfies following requirements (1) to (5): RMS in 1 mm≤λ (space wavelength)≤10 mm is 0.1 nm or more and 3.0 nm or less (1); RMS in 10 μm≤λ (space wavelength)≤1 mm is 0.1 nm or more and 3.0 nm or less (2); RMS in 250 nm≤λ (space wavelength)≤10 μm is 0.05 nm or more and 1.0 nm or less (3); RMS in 100 nm≤λ (space wavelength)≤1 μm is 0.01 nm or more and 1.0 nm or less (4); and RMS in 50 nm≤λ (space wavelength)≤250 nm is 0.01 nm or more and 1.0 nm or less (5).

Description

  The present invention relates to a mirror for ArF lithography and an optical member for ArF lithography suitable for the mirror.

  Conventionally, in an optical lithography technique, an exposure apparatus for manufacturing an integrated circuit by transferring a fine circuit pattern onto a wafer has been widely used. As integrated circuits become highly integrated and highly functional, miniaturization of integrated circuits advances, and the exposure apparatus is required to form a high-resolution circuit pattern on the wafer surface with a deep focal depth. Short wavelength is being promoted. As an exposure light source, an ArF excimer laser (wavelength 193 nm) is going to be used, proceeding from conventional g-line (wavelength 436 nm), i-line (wavelength 365 nm) or KrF excimer laser (wavelength 248 nm).

Conventionally, a refractive optical system has been studied as an optical system for ArF lithography using an ArF excimer laser as an exposure light source.
However, when a refractive optical system is used for ArF lithography, (1) imaging unevenness occurs due to temperature rise during light transmission, and (2) the optical member cannot be increased in curvature due to the numerical aperture (NA). However, there is a problem that it is difficult to produce an optical member having a large size with a uniform composition, and therefore the use of a reflective optical system is being studied (see Patent Document 1).

  However, since a refractive optical system has been conventionally studied as an optical system for ArF lithography, an optical member having characteristics suitable for a reflective optical system for ArF lithography, specifically, a mirror for ArF lithography. There were no optical members with suitable properties.

JP 2005-017543 A

  In order to solve the above-mentioned problems of the prior art, an object of the present invention is to provide an optical member having characteristics suitable for an ArF lithography mirror, and an ArF lithography mirror using the optical member.

In order to achieve the above object, the present invention provides an optical member for ArF lithography having an optical surface satisfying the following (1) to (5) and having a coefficient of thermal expansion (CTE) at 22 ° C. of 0 ± 200 ppb / ° C.
RMS at 1 mm ≦ λ (spatial wavelength) ≦ 10 mm is 0.1 nm to 3.0 nm (1)
RMS at 10 μm ≦ λ (spatial wavelength) ≦ 1 mm is 0.1 nm to 3.0 nm (2)
RMS at 250 nm ≦ λ (spatial wavelength) ≦ 10 μm is 0.05 nm or more and 1.0 nm or less (3)
RMS at 100 nm ≦ λ (spatial wavelength) ≦ 1 μm is 0.01 nm or more and 1.0 nm or less (4)
RMS at 50 nm ≦ λ (spatial wavelength) ≦ 250 nm is 0.01 nm to 1.0 nm (5)

The optical member for ArF lithography of the present invention is preferably a silica glass containing at least one selected from the group consisting of TiO ₂ , SnO ₂ , ZrO ₂ , HfO ₂ , N and F as a dopant.

The optical member for ArF lithography of the present invention has an H ₂ concentration of less than 5 × 10 ¹⁷ molecules / cm ³ , an OH group concentration of less than 700 wtppm, and an O ₂ concentration of 5 × 10 ¹⁷ molecules / cm 3. Preferably it is less than ³ .

  The optical member for ArF lithography of the present invention preferably has a specific elastic modulus of 29.0 MNm / kg or more.

  The optical member for ArF lithography of the present invention preferably has a Young's modulus of 65 GPa or more.

  The optical member for ArF lithography of the present invention preferably has a thermal conductivity of 1.2 W / (m · ° C.) or higher at 22 ° C. of the optical member.

  The optical member for ArF lithography of the present invention preferably has a Vickers hardness Hv of the optical surface of the optical member of 6.5 GPa or more.

  The optical member for ArF lithography of the present invention preferably has a virtual temperature Tf of 900 to 1100 ° C.

  In the present invention, an aluminum thin film is formed as a reflective film on the optical surface of the mirror optical member for ArF lithography of the present invention, and at least one dielectric thin film is formed as a protective film on the reflective film. A mirror for ArF lithography is provided.

  The ArF lithography mirror of the present invention preferably has a light reflectance at a wavelength of 193 nm of 80% or more.

In the mirror for ArF lithography of the present invention, when the ArF laser beam is irradiated with 10 ⁷ shots at 20 mJ / cm ² / pulse, the decrease in the light reflectance at a wavelength of 193 nm before and after the ArF laser beam irradiation is 2.0% or less. Preferably there is.

The mirror for ArF lithography of the present invention has a very high light reflectivity in the wavelength region (193 nm) of ArF laser light, a very low decrease in light reflectivity in the wavelength region due to ArF laser light irradiation, and uneven imaging. Is suitable as a mirror for ArF lithography.
The optical member for ArF lithography of the present invention has characteristics suitable for a mirror for ArF lithography.

The present invention will be described below.
The optical member for ArF lithography of the present invention has a thermal expansion characteristic suitable for a mirror for ArF lithography. Specifically, the optical member for ArF lithography of the present invention has a coefficient of thermal expansion (CTE) at 22 ° C. of 0 ± 200 ppb / ° C.
When performing ArF lithography, the temperature in the exposure apparatus is strictly controlled to be 22 ° C. for the purpose of preventing dimensional changes due to temperature changes of the mirrors in the exposure apparatus.
If the thermal expansion coefficient (CTE) at 22 ° C. of the optical member for ArF lithography used for the mirror is 0 ± 200 ppb / ° C., the thermal expansion coefficient of the optical member at the time of ArF lithography is almost zero. It is possible to prevent the inner mirror from changing its dimensions due to temperature changes.
The optical member for ArF lithography of the present invention preferably has a coefficient of thermal expansion (CTE) at 22 ° C. of 0 ± 6 ppb / ° C., more preferably 0 ± 5 ppb / ° C., and 0 ± 4 ppb / ° C. More preferably.

As an optical member having a coefficient of thermal expansion (CTE) satisfying the above range at 22 ° C., silica glass containing at least one selected from the group consisting of TiO ₂ , SnO ₂ , ZrO ₂ , HfO ₂ , N and F as a dopant. Is mentioned.

The dopant content in silica glass varies depending on the type of dopant. When TiO ₂ is included as a dopant, the TiO ₂ content is preferably 3 to 9% by mass, more preferably 5 to 8.5% by mass, and further preferably 6 to 7.5% by mass. preferable.
When SnO ₂ is contained as a dopant, the SnO ₂ content is preferably 0.001 to 5% by mass, more preferably 0.01 to 3% by mass, and 0.1 to 1.5% by mass. More preferably it is.
When ZrO ₂ is contained as a dopant, the ZrO ₂ content is preferably 0.001 to 5% by mass, more preferably 0.01 to 3% by mass, and 0.1 to 1% by mass. Is more preferable.
When HfO ₂ is included as a dopant, the HfO ₂ content is preferably 0.001 to 5% by mass, more preferably 0.01 to 3% by mass, and 0.1 to 1% by mass. Is more preferable.
When N is contained as a dopant, the N content is preferably 0.001 to 5% by mass, more preferably 0.01 to 3% by mass, and further preferably 0.1 to 2% by mass. preferable.
When F is contained as the dopant, the F content is preferably 0.01 to 3% by mass, more preferably 0.1 to 2% by mass, and further preferably 0.5 to 1% by mass. preferable.
When F is contained as a dopant, it is usually preferable to contain it together with another dopant, for example, TiO ₂ . In this case, the total content of both is preferably 4 to 9% by mass, more preferably 5 to 8.5% by mass, and even more preferably 6 to 8% by mass.

Among these, silica glass (TiO ₂ —SiO ₂ glass) containing TiO ₂ as a dopant in the above content is easily vaporized in a glass manufacturing method described later, and it is easy to obtain homogeneous glass. Preferred for reasons.

The optical member for ArF lithography of the present invention has surface characteristics suitable as a mirror for ArF lithography. Specifically, the optical surface of the optical member for ArF lithography of the present invention satisfies the following (1) to (5).
RMS at 1 mm ≦ λ (spatial wavelength) ≦ 10 mm is 0.1 nm to 3.0 nm (1)
RMS at 10 μm ≦ λ (spatial wavelength) ≦ 1 mm is 0.1 nm to 3.0 nm (2)
RMS at 250 nm ≦ λ (spatial wavelength) ≦ 10 μm is 0.05 nm or more and 1.0 nm or less (3)
RMS at 100 nm ≦ λ (spatial wavelength) ≦ 1 μm is 0.01 nm or more and 1.0 nm or less (4)
RMS at 50 nm ≦ λ (spatial wavelength) ≦ 250 nm is 0.01 nm to 1.0 nm (5)
In the above (3) to (5), the surface roughness of the 1 μm □ region, 2 μm □ region, and 10 μm □ region is measured with an atomic force microscope (AFM), and a bandpass filter that becomes a predetermined spatial region is obtained. It can be calculated from the multiplied result. The above (2) is calculated from the result obtained by measuring the surface roughness of a 2 mm □ area with a non-contact surface shape measuring instrument (for example, NewView manufactured by ZYGO) and applying a bandpass filter that becomes a predetermined spatial area. Can do. The above (1) measures the surface roughness of a 30-40 mm wide line-shaped area with a non-contact surface shape measuring instrument (for example, NewView manufactured by ZYGO), and applies a bandpass filter that becomes a predetermined spatial area. It can be calculated from the results.
In addition, the means for making an optical surface satisfy | fill said (1)-(5) is mentioned later.

If the RMS at each λ (spatial wavelength) is in the above range, the surface roughness of the optical surface becomes an appropriate size, and the following effects are expected to be obtained.
Suppression of particle contamination When creating a mirror using the optical member, or when performing ArF lithography using the created mirror, the optical member is charged to cause particles to adhere or cause particle generation. Become a problem. Usually, charging of such an optical member is contact peeling charging and is known to depend on a contact area between the holding surface and another material such as a metal. Therefore, for example, when the surface roughness of the optical member is reasonably large, the contact area between the optical member and another material such as a metal is reduced, and the charge amount of the optical member is also reduced. As a result, it is considered that particles adhering to static electricity can be suppressed, and similarly, adhesion of particles in the liquid can be suppressed, so that the cleaning effect of the optical member is considered to increase.
Improvement of film adhesion When forming an aluminum thin film and a dielectric thin film on the optical surface of the optical member for ArF lithography in the procedure described later, if the surface roughness of the optical surface of the optical member is reasonably large, these thin films It is possible to improve the adhesion.
Promotion of heat transfer from the multilayer reflective film to the optical member When the surface roughness of the optical surface of the optical member for ArF lithography is reasonably large, the surface area of the optical surface is increased. When the dielectric thin film is formed, heat conduction from the thin film to the optical member is promoted.
Suppression of transmission of stray light from the side surface When stray light enters the side surface (with respect to the optical surface and the holding surface) of the optical member during the ArF lithography, the intruding light repeatedly reflects off the surface of the optical member and then enters the outside. Released. At this time, if the surface roughness of the optical surface is reasonably large, incident light is scattered by the optical surface, so that straightness is impaired. As a result, a part of the light enters the holding surface of the optical member and is absorbed by the electrostatic chuck (a metal body constituting the electrostatic chuck or a metal film formed on the electrostatic chuck) in contact with the holding surface. The As a result, light emitted as stray light from the inside of the optical member is reduced.

In the optical member for ArF lithography of the present invention, the optical surface preferably satisfies the following (6) to (10).
RMS at 1 mm ≦ λ (space wavelength) ≦ 10 mm is 0.3 nm or more and 2.0 nm or less (6)
RMS at 10 μm ≦ λ (spatial wavelength) ≦ 1 mm is 0.3 nm or more and 2.0 nm or less (7)
RMS at 250 nm ≦ λ (spatial wavelength) ≦ 10 μm is 0.1 nm or more and 0.5 nm or less (8)
RMS at 100 nm ≦ λ (spatial wavelength) ≦ 1 μm is 0.01 nm or more and 0.1 nm or less (9)
RMS at 50 nm ≦ λ (spatial wavelength) ≦ 250 nm is 0.01 nm or more and 0.1 nm or less (10)

  In the optical member for ArF lithography according to the present invention, it is preferable that the amount of gas released from the optical member is extremely small when performing ArF lithography since the pattern accuracy during ArF lithography will not be adversely affected.

In order to set the amount of gas released from the optical member within the above range, the H ₂ concentration in the optical member is preferably less than 5 × 10 ¹⁷ molecules / cm ³ and is preferably 1 × 10 ¹⁷ molecules / cm. More preferably, it is less than ³ , more preferably less than 5 × 10 ¹⁶ molecules / cm ³ , and particularly preferably less than 1 × 10 ¹⁶ molecules / cm ³ .
The OH group concentration in the optical member is preferably less than 700 wtppm, more preferably less than 200 mass ppm, further preferably less than 100 wtppm, and particularly preferably less than 50 wtppm.
In addition, the O ₂ concentration in the optical member is preferably less than 5 × 10 ¹⁷ molecules / cm ³ , more preferably less than 1 × 10 ¹⁷ molecules / cm ³ , and 5 × 10 ¹⁶ molecules / cm ^3. More preferably, it is more preferably less than 1 × 10 ¹⁶ molecules / cm ³ .

The H ₂ concentration of the optical member can be measured by the following procedure using a thermal desorption spectrometer (TDS) manufactured by Electronic Science Co., Ltd. based on the specification of Japanese Patent No. 3298974.
An optical member (measurement sample) to be measured is placed in a thermal desorption analyzer, the inside of the measurement chamber is evacuated to a predetermined vacuum level, the optical member is heated, and the mass number of the generated gas is analyzed. measured by installed device inside the mass spectrometer, (plotted in relation to temperature and a detected amount of H ₂₎ Atsushi Nobori profile of H ₂ to create. Next, a temperature-programmed desorption profile of H ₂ is created for a standard sample made of the same material as the optical member and having a known H ₂ concentration. Based on the Atsushi Nobori profile of H ₂ created, measurement samples by ratio of integrated intensities of the desorption peak of H ₂ and the measurement sample and the standard sample, and calculates the number of desorbed from the measurement sample H ₂ H ₂ concentration contained in can be obtained.

  The OH group concentration of the optical member can be measured as follows. The absorption spectrum of the optical member is measured from the measurement with an infrared spectrophotometer, and the OH group concentration is determined from the absorption peak at a wavelength of 2.7 μm of the spectrum. The detection limit by this method is usually about 0.1 wtppm.

The method for measuring the O ₂ concentration of the optical member is as follows. Excitation is performed with a laser having a wavelength of 1064 nm or 765 nm, and emission of a 1272 nm peak is measured. The measurement is carried out using a detector capable of measuring light having a wavelength of 1272nm (L. Skuja and B. Guttler, "Detection of Interstitial Oxygen Molecules in SiO 2 Glass by a Direct Photoexcitation of the Infrared Luminescence of Singlet O 2" Physical Review Letters, (USA), 1996, Vol. 77, No. 10, P.2093-2096)).
Since the O ₂ concentration is proportional to the peak intensity I of the emission spectrum, the average O ₂ concentration can be calculated from the ratio with the emission peak intensity of a standard sample whose O ₂ concentration is known in advance. When there is no standard sample, the Raman intensity peak intensity I _{R of} Raman shift = 490 cm ⁻¹ is constant regardless of the sample, and the ratio I of the emission spectrum peak intensity I and the Raman peak intensity I _R From / I _R , the average O ₂ concentration can be calculated by the relational expression of average O ₂ concentration≈5 × 10 ¹⁷ I / I _R [cm ⁻³ ].

  The optical member for ArF lithography of the present invention has a specific elastic modulus of 29.0 MNm / kg or more when forming an aluminum thin film and a dielectric thin film on the optical surface of the optical member for ArF lithography according to the procedure described later. It is preferable because the amount of deformation of the optical member is reduced by holding it with an electrostatic chuck during lithography. The optical member for ArF lithography of the present invention preferably has a specific modulus of 30.0 MNm / kg or more, more preferably 30.5 MNm / kg or more, and 31.0 MNm / kg or more. Particularly preferred.

  The optical member for ArF lithography of the present invention preferably has a Young's modulus of 65 GPa or more in order to prevent deformation due to film stress of the aluminum thin film and the dielectric thin film formed on the optical surface. In the optical member for ArF lithography of the present invention, the Young's modulus is more preferably 67 GPa or more, further preferably 67.5 GPa or more, and particularly preferably 68 GPa or more.

  Since ArF lithography is performed in a nitrogen atmosphere, it is preferable that the optical member for ArF lithography of the present invention has a high thermal conductivity in order to prevent a temperature change of the optical member during ArF excimer laser irradiation. Specifically, the thermal conductivity of the optical member at 22 ° C. is preferably 1.2 W / (m × ° C.) or more, more preferably 1.3 W / (m × ° C.) or more, More preferably, it is at least 35 W / (m × ° C.).

  If the optical surface has a low Vickers hardness Hv, pit-like defects (that is, concave defects) are likely to occur when the optical surface is polished. Therefore, the optical member for ArF lithography of the present invention has an optical surface with a Vickers hardness Hv of High is preferred. Specifically, the Vickers hardness Hv is preferably 6.5 GPa or more, more preferably 6.7 GPa or more, and further preferably 6.8 GPa or more.

When the optical member for ArF lithography of the present invention is composed of silica glass containing at least one selected from the group consisting of TiO ₂ , SnO ₂ , ZrO ₂ , HfO ₂ , N and F as a dopant, the Vickers of the optical surface In order to increase the hardness Hv, the fictive temperature Tf of the silica glass is preferably high. Specifically, the fictive temperature Tf of the silica glass is preferably 900 ° C. or higher and 1100 ° C. or lower, more preferably 930 ° C. or higher and 1070 ° C. or lower, and further preferably 950 ° C. or higher and 1050 ° C. or lower. .
The fictive temperature of silica glass containing at least one selected from the group consisting of TiO ₂ , SnO ₂ , ZrO ₂ , HfO ₂ , N and F as a dopant can be measured by a known procedure. In Examples described later, the fictive temperature of silica glass containing TiO ₂ as a dopant was measured by the following procedure.

For the mirror-polished silica glass, an absorption spectrum is obtained using an infrared spectrometer (in the examples described below, Magna 760 manufactured by Nikolet is used). At this time, the data interval is set to about 0.5 cm ⁻¹ , and the average value obtained by scanning 64 times is used as the absorption spectrum. In the infrared absorption spectrum thus obtained, the peak observed in the vicinity of about 2260 cm ⁻¹ is due to the overtone of stretching vibration due to the Si—O—Si bond of silica glass. Using this peak position, a calibration curve is created with glass having the same fictive temperature and the same composition, and the fictive temperature is obtained. Alternatively, the reflection spectrum of the surface is similarly measured using a similar infrared spectrometer. In the infrared reflection spectrum thus obtained, the peak observed in the vicinity of about 1120 cm ⁻¹ is due to stretching vibration due to the Si—O—Si bond of silica glass. Using this peak position, a calibration curve is created with glass having the same fictive temperature and the same composition, and the fictive temperature is obtained. Note that the shift of the peak position due to the change in the glass composition can be extrapolated from the composition dependency of the calibration curve.

When the optical member for ArF lithography is composed of silica glass containing at least one selected from the group consisting of TiO ₂ , SnO ₂ , ZrO ₂ , HfO ₂ , N and F as a dopant, the silica glass has a fictive temperature Is preferably within 50 ° C., more preferably within 30 ° C. If the variation in the fictive temperature exceeds the above range, there may be a difference in the coefficient of thermal expansion depending on the location.
In this specification, “variation of virtual temperature” is defined as the difference between the maximum value and the minimum value of virtual temperature within 30 mm × 30 mm in at least one plane.
The variation in fictive temperature can be measured as follows. The silica glass body molded into a predetermined size is sliced into a glass block of 50 mm × 50 mm × 6.35 mm. By measuring the fictive temperature according to the above-mentioned method at intervals of 10 mm on the 50 mm × 50 mm surface of this glass block, the variation in fictive temperature of the molded silica glass body is obtained.

When the optical member for ArF lithography is composed of silica glass containing at least one selected from the group consisting of TiO ₂ , SnO ₂ , ZrO ₂ , HfO ₂ , N and F as a dopant, Making the composition ratio of (for example, TiO ₂ ) and SiO ₂ uniform is important in terms of reducing variations in the thermal expansion coefficient within the silica glass. Specifically, the PV (Peak to Valley) value of the dopant concentration in the optical member is preferably 0.15% by mass or less, and more preferably 0.10% by mass or less.
The PV value of the dopant concentration in the optical member can be determined by the following method. On a line passing through the center point of the surface of the optical member, it is continuously cut from end to end so as to include a surface area of 50 mm × 50 mm, and 50 mm × 50 mm (thickness is arbitrary, for example, thickness 10 mm ) To obtain a plurality of blocks. After grinding and polishing the 50 mm × 50 mm surface of this glass block to a flat surface, composition analysis is performed by fluorescent X-ray analysis, and the PV value of the dopant concentration is determined by determining the variation of the dopant concentration on the line passing through the center point of the surface of the optical member. obtain.
Alternatively, the composition of the similarly cut block is analyzed by a wet analysis method such as ICP emission analysis, and the variation of the dopant concentration on the line passing through the center point of the surface of the optical member is obtained to obtain the PV value of the dopant concentration. obtain.

As described above, the optical member for ArF lithography of the present invention is preferably silica glass (TiO ₂ —SiO ₂ glass) containing TiO ₂ as a dopant in the above content.
As a method for producing TiO ₂ -SiO ₂ glass, method can be employed including the following (a) ~ (e) step.

(A) Process TiO ₂ —SiO ₂ glass fine particles (soot) obtained by flame hydrolysis or thermal decomposition of SiO ₂ precursor and TiO ₂ precursor, which are glass forming raw materials, are deposited and grown on a substrate by a soot method. To form a porous TiO ₂ —SiO ₂ glass body. The soot method includes an MCVD method, an OVD method, and a VAD method, depending on how to make the soot method. Among these, the VAD method is preferable because it is excellent in mass productivity, and a glass having a uniform composition in a large area can be obtained by adjusting manufacturing conditions such as the size of the substrate.

The glass forming raw material is not particularly limited as long as it is a gasifiable raw material. Examples of the SiO ₂ precursor include chlorides such as SiCl ₄ , SiHCl ₃ , SiH ₂ Cl ₂ , and SiH ₃ Cl, SiF ₄ , and SiHF _3. , fluorides such as SiH ₂ F _2, SiBr _4, bromides such as SiHBr _3, halogenated silicon compounds such as iodide such as SiI _{4, R n Si (OR) 4-} n ( here R is 1 carbon atoms 4), n is an integer of 0 to 3. A plurality of Rs may be the same or different from each other), and examples of the TiO ₂ precursor include TiCl ₄ and TiBr _4. halogenated titanium compound and _{R n Ti (OR) 4-} n ( wherein R is an alkyl group having 1 to 4 carbon atoms, n represents an integer. multiple R of 0-3 may be the same or different from each other .) The alkoxy titanium shown by these is mentioned. In addition, Si and Ti compounds such as silicon titanium double alkoxide can be used as the SiO ₂ precursor and the TiO ₂ precursor.
As a base material, a seed rod made of quartz glass (for example, a seed rod described in Japanese Patent Publication No. 63-24973) can be used. Moreover, you may use not only rod shape but a plate-shaped base material.

(B) Step The porous TiO ₂ —SiO ₂ glass body obtained in the step (a) is heated to a densification temperature in an inert gas atmosphere or a reduced pressure atmosphere to obtain a TiO ₂ —SiO ₂ dense body. . In the present invention, the densification temperature refers to a temperature at which the porous glass body can be densified until no voids can be confirmed with an optical microscope, preferably 1250 to 1550 ° C., particularly preferably 1350 to 1450 ° C. As the inert gas, helium is preferable. In such an atmosphere, it is preferable to perform the treatment at a pressure of about 10,000 to 200,000 Pa. In this specification, “Pa” means absolute pressure, not gauge pressure.
In step (b), since the homogeneity of the TiO ₂ —SiO ₂ dense body is increased, the porous TiO ₂ —SiO ₂ glass body is placed under reduced pressure (preferably 13000 Pa or less, particularly 1300 Pa or less), Next, it is preferable to introduce an inert gas into an inert gas atmosphere at a predetermined pressure.
Further, in the step (b), since the homogeneity of the TiO ₂ —SiO ₂ dense body is increased, the porous TiO ₂ —SiO ₂ glass body is at room temperature in an inert gas atmosphere or at a temperature equal to or lower than the densification temperature. After holding, it is preferable to raise the temperature to the densification temperature.
Further, when obtaining a TiO ₂ -SiO ₂ densified body was raised to densification temperature under a reduced pressure atmosphere, the pressure is preferably set to 13,000 Pa, and more preferably to less 1300 Pa.
In addition, after the completion of the step (a) and before entering the step (b), the obtained porous TiO ₂ —SiO ₂ glass body is heat-treated at a temperature range of 1000 ° C. or higher and a densification temperature or lower for a predetermined time or longer. This makes it easier to handle. Thereby, it is possible not only to suppress the generation of cracks in the TiO ₂ —SiO ₂ dense body obtained in the step (b) but also to reduce the concentration of gas dissolved in the TiO ₂ —SiO ₂ dense body. For the purpose of reducing the gas concentration, the heat treatment atmosphere is preferably a reduced pressure.

(C) Step TiO ₂ —SiO ₂ dense body obtained in step (b) is heated to a transparent vitrification temperature to obtain a transparent TiO ₂ —SiO ₂ glass body. In this specification, the transparent vitrification temperature refers to a temperature at which crystals cannot be confirmed with an optical microscope and a transparent glass is obtained, preferably 1350 to 1750 ° C., particularly preferably 1400 to 1700 ° C.
The atmosphere is preferably an atmosphere of 100% inert gas such as helium or argon, or an atmosphere mainly composed of an inert gas such as helium or argon. The pressure may be reduced pressure or normal pressure. In the case of reduced pressure, 13000 Pa or less is preferable.

(D) Step The transparent TiO ₂ —SiO ₂ glass body obtained in the step (c) is put into a mold and heated to a temperature equal to or higher than the softening point to form a desired shape, and the molded TiO ₂ —SiO ₂ glass body is formed. Get. The molding process temperature is preferably 1500 to 1800 ° C. When the temperature is 1500 ° C. or higher, the viscosity is sufficiently lowered to the extent that the transparent TiO ₂ —SiO ₂ glass is substantially deformed by its own weight. Further, the growth of cristobalite, which is a crystal phase of SiO ₂ , or the growth of rutile or anatase, which is a crystal phase of TiO ₂ , hardly occurs, and so-called devitrification can be prevented. Below 1800 ° C., sublimation of SiO ₂ can be suppressed.
The above procedure may be repeated a plurality of times. That is, after the transparent TiO ₂ —SiO ₂ glass body is put into a mold and heated to a temperature above the softening point, the obtained molded body is put into another mold and heated to a temperature above the softening point. You may implement.
The steps (c) and (d) can be performed continuously or simultaneously. When the size of the glass obtained in the step (c) is sufficiently large, the transparent TiO ₂ —SiO ₂ glass body obtained in the step (c) is cut into a predetermined size without performing the step (d). it is can be a formed TiO ₂ -SiO ₂ glass body.

(E) Step After raising the temperature of the transparent TiO ₂ —SiO ₂ glass body obtained in step (c) or the molded TiO ₂ —SiO ₂ glass body obtained in step (d) to a temperature of 1100 ° C. or higher, An annealing process for lowering the temperature to 700 ° C. or lower at an average temperature lowering rate of 100 ° C./hr or lower is performed to control the fictive temperature of the glass. Alternatively, in the temperature lowering process from a temperature of 1100 ° C. or higher in the vitrification process or the molding process, the obtained transparent TiO ₂ —SiO ₂ glass body or molded TiO ₂ —SiO ₂ glass body is 100 ° C./hr from 1100 ° C. to 700 ° C. An annealing process for lowering the temperature at the following average temperature lowering rate is performed to control the virtual temperature of the glass. In these cases, the average temperature lowering rate is more preferably 10 ° C./hr or less, further preferably 5 ° C./hr or less, and particularly preferably 2.5 ° C./hr or less. In addition, after the temperature is lowered to 700 ° C. or lower, it can be allowed to cool. The atmosphere is not particularly limited.

Also, prior to step (e), a glass body after transparent vitrification may be heated over 20 hours at T ₁ + 400 ° C. or higher temperatures ((d ') step). Here, T ₁ is the annealing point (° C.) of the TiO ₂ —SiO ₂ glass body to be produced. Thereby, the streak in the TiO ₂ —SiO ₂ glass body is reduced.
A streak is a non-uniform composition (composition distribution) of a TiO ₂ —SiO ₂ glass body. In the TiO ₂ —SiO ₂ glass body having the strie, there are portions having different TiO ₂ concentrations. Since the site TiO ₂ concentration is high coefficient of linear thermal expansion (CTE) becomes negative, during the annealing process performed in the production process of the TiO ₂ -SiO ₂ glass body, TiO ₂ concentration is high The site tends to swell. At this time, when the TiO ₂ concentration adjacent the TiO ₂ concentration is high sites exist lower portion, so that the compressive stress expansion of the TiO ₂ concentration is high sites can be prevented is applied. As a result, a stress distribution occurs in the TiO ₂ —SiO ₂ glass body. Hereinafter, in the present specification, such a stress distribution is referred to as a “stress distribution caused by a strie”.

If there is a distribution of stress generated by such a streak in a TiO ₂ —SiO ₂ glass body used as an optical member for ArF lithography, a difference occurs in the processing rate when the optical surface of the optical member is finished. The surface smoothness of the optical surface after processing will be affected.
If the above-described step (d ′) is performed, the stress level distribution generated by the striee is used as an ArF reflecting optical system optical member in the TiO ₂ —SiO ₂ glass body manufactured through the subsequent step (e). This is preferable because it is reduced to a level that does not cause a problem.
Specifically, the standard deviation (dev [σ]) of stress generated by the strie is preferably 0.05 MPa or less, more preferably 0.04 MPa or less, and further preferably 0.03 MPa or less. preferable.
Further, the difference (Δσ) between the maximum value and the minimum value of the stress caused by the strie is preferably 0.23 MPa or less, more preferably 0.2 MPa or less, and further preferably 0.15 MPa or less. .
In addition, the stress of a glass body can obtain | require retardation by measuring the area | region about 1 mm x 1 mm using a well-known method, for example, a birefringence microscope, and can obtain | require it from the following formula | equation.
Δ = C × F × n × d
Here, Δ is retardation, C is a photoelastic constant, F is stress, n is a refractive index, and d is a sample thickness.
The stress profile is obtained by the above method, and the standard deviation (dev [σ]) of the stress and the difference (Δσ) between the maximum value and the minimum value of the stress can be obtained therefrom. More specifically, a cube of about 40 mm × 40 mm × 40 mm, for example, is cut out from the transparent TiO ₂ —SiO ₂ glass body, sliced and polished at a thickness of about 1 mm from each surface of the cube, and 30 mm × 30 mm × 0.00 mm. A 5 mm plate glass block is obtained. With a birefringence microscope, helium neon laser light is vertically applied to the 30 mm x 30 mm surface of the glass block, and the magnification is increased to a level at which the striae can be observed sufficiently. The in-plane retardation distribution is examined and converted into a stress distribution. To do. When the striae pitch is fine, it is necessary to reduce the thickness of the sheet glass block to be measured.
In the TiO ₂ —SiO ₂ glass body, at least in the case of the stress measured by the above measurement method, the stress caused by other factors is negligible compared to the stress caused by the strie. Accordingly, the stress obtained by the above method is substantially equal to the stress caused by the strie.

In the step (d ′), the specific procedure is not particularly limited as long as the glass body after transparent vitrification can be heated at a temperature of T ₁ + 400 ° C. or more for 20 hours or more, but the temperature for heating the glass body is not limited. Is too high, foaming and sublimation in the TiO ₂ —SiO ₂ glass body become a problem, which is not preferable. For these reasons, the upper limit of the heating temperature in step (d ′) is preferably less than T ₁ + 600 ° C., more preferably less than T ₁ + 550 ° C., and particularly preferably less than 500 ° C. That, (d ') the heating temperature in the step is preferably less than T ₁ + 400 ° C. or higher T ₁ + 600 ° C., more preferably less than T ₁ + 400 ° C. or higher _{T 1 + 550 ℃, T 1} + 450 ℃ or higher More preferably, it is less than T ₁ + 500 ° C.

(D ') is not particularly restricted upper limit of the heating time is also in the process, when the heating time is too long, no longer does not contribute to the reduction of Sutorie in TiO ₂ -SiO ₂ glass body yields a TiO ₂ -SiO ₂ glass body Is preferably 240 hours or less, and more preferably 150 hours or less, for reasons such as lowering the temperature or increasing the cost due to heating. Also, the lower limit of the heating time is not particularly limited, but when the heating time is too short, it is difficult to obtain the effect of reducing the streaks in the TiO ₂ —SiO ₂ glass body. Therefore, the heating time is preferably more than 24 hours, more preferably more than 48 hours, and particularly preferably more than 96 hours.
The step (d ′) can be performed continuously or simultaneously with the step (e). In addition, the step (d ′) can be performed continuously or simultaneously with the steps (c) and (d).

In order to exclude inclusions such as foreign matters and bubbles from the produced TiO ₂ —SiO ₂ glass, in the production process described above, it is particularly preferable to suppress contamination in the step (a), and further to the steps (b) to (d It is necessary to accurately control the temperature conditions of
In the above, the procedure for producing the TiO ₂ —SiO ₂ glass body by the soot method has been shown, but the present invention is not limited to this, and the TiO ₂ —SiO ₂ glass body can also be produced by a direct method. In this case, in the step (a), a transparent TiO ₂ —SiO ₂ glass is directly obtained by hydrolyzing and oxidizing a silica precursor and a titania precursor as glass forming raw materials in an oxyhydrogen flame at 1800 to 2000 ° C. Get the body. That is, by the step (a), a transparent TiO ₂ —SiO ₂ glass body can be obtained without performing the steps (b) and (c), and the molded TiO ₂ —SiO ₂ glass body is obtained by the step (d). Then, the step (e) may be performed. In addition, the transparent TiO ₂ —SiO ₂ glass body obtained in the step (a) is cut into a predetermined size to obtain a molded TiO ₂ —SiO ₂ glass body, and then the step (e) is performed. Good. The resulting transparent TiO ₂ —SiO ₂ glass body contains H ₂ and OH. At this time, the OH concentration of the transparent TiO ₂ —SiO ₂ glass body can be adjusted by adjusting the flame temperature and gas concentration. Furthermore, when the glass produced by the direct method is in a vacuum, in a reduced pressure atmosphere or at normal pressure, in an atmosphere having an H ₂ concentration of 1000 ppm by volume or less and an O ₂ concentration of 18% by volume or less, 700 ° C. to 1800 ° C. A method of degassing by holding at a temperature for 10 minutes to 90 days can also be employed.

In order for the optical surface of the optical member for ArF lithography to satisfy the following (1) to (5), the optical surface may be subjected to a three-stage mechanical polishing process described below.
RMS at 1 mm ≦ λ (spatial wavelength) ≦ 10 mm is 0.1 nm to 3.0 nm (1)
RMS at 10 μm ≦ λ (spatial wavelength) ≦ 1 mm is 0.1 nm to 3.0 nm (2)
RMS at 250 nm ≦ λ (spatial wavelength) ≦ 10 μm is 0.05 nm or more and 1.0 nm or less (3)
RMS at 100 nm ≦ λ (spatial wavelength) ≦ 1 μm is 0.01 nm or more and 1.0 nm or less (4)
RMS at 50 nm ≦ λ (spatial wavelength) ≦ 250 nm is 0.01 nm to 1.0 nm (5)

[First polishing step]
The first polishing step is performed for the purpose of giving a rough shape of the optical surface, and mainly contributes to the achievement of the above (1). The target value of the surface roughness of the optical surface after the first polishing step is 2 nm in terms of surface roughness RMS when a 10 μm square region is measured with an atomic force microscope (AFM).
When the application of the optical member for ArF lithography is a mirror for ArF lithography, dressing is performed using a polishing pad in the first polishing step to form a predetermined curved surface, and then a predetermined optical surface is obtained. At this time, as the abrasive, those containing cerium oxide-based abrasive grains are preferably used. The average grain size of the abrasive grains is preferably 500 to 2000 nm, and more preferably 1000 to 1500 nm. .
As the polishing pad, it is preferable to use a hard polishing pad in which polyurethane is bonded to a nonwoven fabric using polyester fibers.

[Second polishing step]
In the second polishing step, the optical surface after the first polishing step is mechanically polished so that the surface roughness RMS is 0.4 nm or less when a 10 μm square region is measured with an atomic force microscope (AFM). It is a process and contributes mainly to the achievement of the above (2) to (4).
As the abrasive, those containing cerium oxide-based abrasive grains are preferably used. The average particle diameter of the abrasive is preferably 100 to 1500 nm, and more preferably 500 to 1000 nm.
Increasing the circulating flow rate of the abrasive is preferable because it contributes to the achievement of the above (2) to (4). Specifically, it is preferably carried out at a circulating flow rate of 1 to 20 L / min, and 2 to 15 L / min. More preferably.
As the polishing pad, a non-woven fabric bonded with polyurethane or a suede type polishing pad in which a NAP layer is formed on a PET sheet is used. The polishing pad has high hardness, specifically, Japan Rubber Association Standard (SRIS) Asker C (ASKER C) is preferably 55 or more.

[Third polishing step]
In the third polishing step, the optical surface after the second polishing step is mechanically polished so that the surface roughness RMS when the 10 μm square region is measured with an atomic force microscope (AFM) is 0.2 nm or less. It is a process and contributes mainly to achievement of said (5).
As an abrasive | polishing agent, what contains the abrasive grain of colloidal silica is used preferably, but it is preferable that the average particle diameter of an abrasive grain is 100 nm or less.
Increasing the circulating flow rate of the abrasive is preferable because it contributes to the achievement of the above (5). Specifically, it is preferably performed at a circulating flow rate of 1 to 20 L / min, and more preferably 2 to 15 L / min. More preferred.
As the polishing pad, a suede polishing pad in which a NAP layer is formed on a nonwoven fabric or PET sheet bonded with polyurethane is preferably used.

  In addition to the above-described three-stage polishing process, a polishing process using a dry etching process such as a gas cluster ion beam as described in Japanese Patent Application Laid-Open No. 2006-240977 and Japanese Patent Application Laid-Open No. 2007-22903 is performed. Since it contributes to achievement of said (1), it is preferable to perform between 2 grinding | polishing processes and 3rd grinding | polishing processes, or after implementation of a 3rd grinding | polishing process. In addition, when the polishing process described in this paragraph is performed after the third polishing process is performed, it is preferable to perform the third polishing process again after the polishing process is performed.

In the mirror for ArF lithography of the present invention, an aluminum thin film is formed as a reflective film on the optical surface of the optical member for the mirror for ArF lithography of the present invention, and at least one dielectric thin film as a protective film is formed on the reflective film. Is formed.
Aluminum is a material having an excellent reflectance in the wavelength range (193 nm) of ArF laser light. However, aluminum is a very active metal, and the surface is quickly oxidized in air to produce aluminum oxide. When the aluminum thin film is left in the air, the light reflectivity, particularly the light reflectivity in the wavelength region (193 nm) of ArF laser light, is abruptly lowered due to light absorption by the aluminum oxide generated on the surface. This tendency is particularly remarkable when ArF laser light is irradiated. This is because the oxidation of the aluminum thin film proceeds particularly rapidly during irradiation with ArF laser light. For this reason, a protective film is formed on the aluminum thin film for the purpose of suppressing oxidation of the aluminum thin film. By forming the protective film, diffusion of oxygen and moisture in the air is suppressed in the aluminum thin film, and oxidation of the aluminum thin film is suppressed.

  The aluminum thin film that forms the reflective film can be formed by any method such as vacuum deposition, ion beam assisted deposition, sputtering, or CVD. Among these, the most suitable method for forming the aluminum thin film is a vacuum deposition method in which the optical member is not heated. This is because according to the vacuum deposition method in which the optical member is not heated, the aluminum thin film is not oxidized during the film formation, and a high reflectance is obtained. Although it is possible to form an aluminum thin film by other methods, in that case, there is a possibility that the mirror has a relatively low reflectance.

As a material for the dielectric thin film that forms the protective film, a known fluoride material that is transparent to ultraviolet rays can be used without particular limitation. Specific examples include yttrium fluoride (YF ₃ ), lanthanum fluoride (LaF ₃ ), gadolinium fluoride (GdF ₃ ), neodymium fluoride (NdF ₃ ), dysprosium fluoride (DyF ₃ ), Magnesium fluoride (MgF ₂ ), aluminum fluoride (AlF ₃ ), lithium fluoride (LiF), cryolite (Na ₃ AlF ₆ ), thiolite (Na ₅ Al ₃ F ₁₄ ) and the like. Among these, yttrium fluoride and aluminum fluoride are materials for constituting the dielectric thin film because the detailed reason is unknown, but a dense and high packing density structure is easily obtained and has appropriate optical characteristics. As the most preferred.

Two or more dielectric thin films forming the protective film may be laminated. For example, the dielectric thin film is an alternating multilayer film composed of a high refractive index layer and a low refractive index layer, and the refractive index and the film thickness are optimized to control the incident angle dependence and polarization dependence of the reflectance. be able to.
Among the above-mentioned fluoride materials, high refractive index substances include yttrium fluoride (YF ₃ ), lanthanum fluoride (LaF ₃ ), gadolinium fluoride (GdF ₃ ), neodymium fluoride (NdF ₃ ), and fluoride display. Sium (DyF ₃ ) is exemplified. On the other hand, as a low refractive index substance, magnesium fluoride (MgF ₂ ), aluminum fluoride (AlF ₃ ), lithium fluoride (LiF), cryolite (Na ₃ AlF ₆ ), thiolite (Na ₅ Al ₃ F ₁₄ ) Is mentioned.
When two or more dielectric thin films are stacked, the total number of dielectric thin films is not particularly limited, and the number of layers may be designed according to optical characteristics such as reflectance and angular characteristics required for the mirror.

It is preferable that the filling density of the dielectric thin film is 90% or more because the effect of suppressing the diffusion of oxygen or water is particularly excellent, and the effect of suppressing the oxidation of the aluminum thin film during ArF laser irradiation is particularly excellent. . When two or more dielectric thin films are stacked, it is sufficient that the packing density of at least one layer is 90% or more.
In this specification, the packing density refers to a value in which the density of the single crystal of the material constituting the thin film is 100% and the density of the actual thin film is expressed as a relative value. As a method for measuring the density of the thin film, a method using X-ray reflectivity measurement is simple. Since the refractive index of a substance with respect to X-rays is smaller than 1, total reflection occurs when X-rays are incident on the thin film at an angle ψ shallower than the critical angle ψc. At this time, if the refractive index of the thin film is n,
n = 1-δ
ψc = (2δ) ^1/2
The refractive index can be obtained from the critical angle measurement. Further, if the composition of the thin film is known, the density can be obtained from the refractive index. When the sample is a multilayer film, a reflectance-incidence angle curve is measured with respect to an incident angle deeper than the critical angle. Interference fringes including information on the thickness, refractive index, and surface roughness of each layer appear in the reflectance-incidence angle curve, and the refractive index of each layer is calculated by curve fitting using these as parameters. If the refractive index is obtained, the density can be calculated from the composition of each layer as described above.

  The dielectric thin film that forms the protective film can be formed by any method such as vacuum deposition, ion beam assisted deposition, sputtering, or CVD. As a method for forming a dielectric thin film having a filling density of 90% or more, it is preferable to use any one of a vacuum evaporation method, an ion beam assisted evaporation method, and a sputtering method performed at a film formation temperature of 150 ° C. or more. . This is because according to these methods, the packing density of the dielectric thin film can be increased, and a packing density of 90% or more can be achieved.

  A general apparatus such as a resistance heating type or an electron beam heating type can be used as it is for a vacuum vapor deposition apparatus used for forming an aluminum thin film or a dielectric thin film, and no special mechanism is required.

When forming an aluminum thin film or a dielectric thin film using an ion beam assisted vapor deposition method, an ion beam assisted vapor deposition apparatus in which an ion source for emitting an ion beam is added to a general vapor deposition apparatus may be used. As the ion source, an arbitrary type such as a Kaufman type or an end hole type can be used. In general, the ion current density in ion beam assisted deposition is often 0 to 10 mA / cm ² , and the acceleration voltage is often about 10 to 100 eV. When an aluminum thin film or a dielectric thin film is formed using a sputtering method, any type of sputtering apparatus such as direct current sputtering, high frequency sputtering, and reactive sputtering can be used.

The mirror for ArF lithography of the present invention has a light reflectance at a wavelength of 193 nm of 80% or more, preferably 85% or more, more preferably 90% or more.
The light reflectivity here refers to the light reflectivity when irradiated with a light beam having an incident angle of 10 degrees and a wavelength of 193 nm, specifically, ArF laser light.

The mirror for ArF lithography of the present invention is excellent in the effect of suppressing the oxidation of the aluminum thin film, and there is little decrease in light reflectivity due to irradiation with ArF laser light. This tendency is particularly true when the packing density of at least one dielectric thin film is 90% or more.
Specifically, when the ArF laser light is irradiated onto the surface of the 10 ⁷ shot multilayer reflective film at 20 mJ / cm ² / pulse, the decrease in the light reflectance at a wavelength of 193 nm before and after the ArF laser light irradiation is 2% or less. Yes, preferably 1% or less, more preferably 0.5% or less.

EXAMPLES Hereinafter, although an Example demonstrates this invention further in detail, this invention is not limited to this.
Example 1 is an example.
[Example 1]
TiCl ₄ and SiCl ₄ as glass-forming raw material for TiO ₂ -SiO ₂ glass, was mixed after each is gasified, TiO ₂ -SiO to subjecting the mixture to heat hydrolysis in an oxyhydrogen flame (flame hydrolysis) ₂ Glass fine particles were deposited and grown on a substrate to form a porous TiO ₂ —SiO ₂ glass body (step (a)).
Since the obtained porous TiO ₂ —SiO ₂ glass body was difficult to handle as it was, it was kept in the atmosphere at 1190 ° C. for 3 hours while being deposited on the substrate, and then removed from the substrate.
Then, heating is performed under reduced pressure, and after heating from 1100 ° C. to 1300 ° C. over 3 hours,
The temperature was raised from 1300 ° C. to 1360 ° C. in 12 hours and held at 1360 ° C. for 2 hours to obtain a dense TiO ₂ —SiO ₂ (step (b)).
The obtained TiO ₂ —SiO ₂ dense body was put in a carbon mold and held at 1680 ° C. for 4 hours to obtain a transparent TiO ₂ —SiO ₂ glass body (step (c)).
The obtained transparent TiO ₂ —SiO ₂ glass body was put in a carbon mold and molded by holding at 1700 ° C. for 4 hours to obtain a molded TiO ₂ —SiO ₂ glass body (step (d)). The obtained molded TiO ₂ —SiO ₂ glass body was held at 1590 ° C. for 120 hours (step (d ′)), cooled to 1000 ° C. at 10 ° C./hr, then held at 1000 ° C. for 3 hours, and 950 After cooling to 10 ° C./hr to 90 ° C., hold at 950 ° C. for 72 hours, cool to 900 ° C. at 5 ° C./hr, hold at 900 ° C. for 72 hours, cool to room temperature, and form TiO ₂ —SiO ₂ body Was obtained (step (e)). In addition, the annealing point T ₁ of the molded TiO ₂ —SiO ₂ body was 1100 ° C.
The obtained molded TiO ₂ —SiO ₂ body is cut into a plate shape having a length of about 153.0 mm × width of about 153.0 mm × thickness of about 6.75 mm using an inner peripheral slicer to produce a plate material. Next, chamfering is performed to obtain a plate material having a length of about 153.0 mm × width of about 153.0 mm × thickness of about 6.7 mm. Using a 20B double-sided lapping machine (manufactured by Speed Fam Co., Ltd.), a slurry obtained by suspending 18 to 20% by mass of AZ # 1000 (trade name, manufactured by Heisei Sankei Co., Ltd.) consisting essentially of Al ₂ O ₃ as an abrasive. Is used to grind the main surface of the plate (the surface on which the multilayer film or absorption layer is formed) until the thickness is about 6.5 mm. Thereafter, the end surface is mirror-finished.
Next, as a first polishing step, a 20B double-side polish machine is used to polish about 50 μm using a foamed polyurethane polishing pad and a polishing agent mainly composed of cerium oxide.
Furthermore, as a second polishing step, a 24B double-sided polisher is used, and a suede type polishing pad in which a NAP layer is formed on a nonwoven fabric bonded with polyurethane. Japan Rubber Association Standard (SRIS) Asker C (ASKER C) Polishing is performed by about 15 μm using a polishing pad having a thickness of 68 and a polishing agent mainly composed of cerium oxide.
Further, the third polishing step is performed with another polishing machine. In this third polishing step, a suede polishing pad in which a NAP layer is formed on a PET sheet and colloidal silica are used.
Thereafter, the flatness measurement is performed, and the shape is corrected by the gas cluster ion beam by an amount corresponding to the final flatness in consideration of the flatness change in polishing in the next step.
Finally, the polished surface is finished using a suede type polishing pad in which a NAP layer is formed on a PET sheet and colloidal silica.

[Example 2] (Comparative example)
In Example 1, the molded TiO ₂ —SiO ₂ glass body obtained in the step (d) was polished as it was. In the second polishing step, a polishing pad in which the Japan Rubber Association Standard (SRIS) Asker C was 55 was used. The other methods were the same as in Example 1.

The surface shape measurement result of the produced glass is as follows.
RMS at 1mm ≦ λ ≦ 10mm
Measuring method: ZYGO NewView
Example 1: 0.9nm
Example 2: 3.6 nm
RMS at 10μm ≦ λ ≦ 1mm
Measuring method: ZYGO NewView
Example 1: 0.43 nm
Example 2: 3.51 nm
RMS at 250 nm ≦ λ ≦ 10 μm
Measuring method: AFM
Example 1: 0.39 nm
Example 2: 1.20 nm
RMS at 100 nm ≦ λ ≦ 1 μm
Measuring method: AFM
Example 1: 0.023 nm
Example 2: 0.081 nm
RMS at 50 nm ≦ λ ≦ 250 nm
Measuring method: AFM
Example 1: 0.027 nm
Example 2: 0.052 nm

Table 1 shows the physical properties of the produced glass. The TiO ₂ concentration is about 7.1% by mass, and the PV value of the dopant concentration is 0.15% by mass or less. The variation in the fictive temperature of the optical member of Example 1 is 10 ° C.

A film is formed on the optical member obtained in Example 1 by the following method to create a mirror. An optical member is installed in the vapor deposition apparatus, and aluminum is deposited to a thickness of 150 nm by a vapor deposition method. Next, the optical member is heated in vacuum to 200 ° C. to form a magnesium fluoride film having a thickness of 40 nm. The packing density of the obtained film is 92%. With respect to the mirror thus obtained, the light reflectance at a wavelength of 193 nm before and after irradiation with ArF laser light when irradiated with 10 ⁷ shot ArF laser light at 20 mJ / cm ² / pulse is 82%, respectively. There is no decline in the rate.

Claims (11)

An optical member for ArF lithography having an optical surface satisfying the following (1) to (5) and having a coefficient of thermal expansion (CTE) at 22 ° C. of 0 ± 200 ppb / ° C.
RMS at 1 mm ≦ λ (spatial wavelength) ≦ 10 mm is 0.1 nm to 3.0 nm (1)
RMS at 10 μm ≦ λ (spatial wavelength) ≦ 1 mm is 0.1 nm to 3.0 nm (2)
RMS at 250 nm ≦ λ (spatial wavelength) ≦ 10 μm is 0.05 nm or more and 1.0 nm or less (3)
RMS at 100 nm ≦ λ (spatial wavelength) ≦ 1 μm is 0.01 nm or more and 1.0 nm or less (4)
RMS at 50 nm ≦ λ (spatial wavelength) ≦ 250 nm is 0.01 nm to 1.0 nm (5) The optical member for ArF lithography according to claim 1, wherein the optical member is silica glass containing at least one selected from the group consisting of TiO ₂ , SnO ₂ , ZrO ₂ , HfO ₂ , N, and F as a dopant. The H ₂ concentration in the optical member is less than 5 × 10 ¹⁷ molecules / cm ³ , the OH group concentration is less than 700 wtppm, and the O ₂ concentration is less than 5 × 10 ¹⁷ molecules / cm ^3. The optical member for ArF lithography as described.   The optical member for ArF lithography according to any one of claims 1 to 3, wherein the specific elastic modulus of the optical member is 29.0 MNm / kg or more.   The optical member for ArF lithography according to any one of claims 1 to 4, wherein the optical member has a Young's modulus of 65 GPa or more.   The optical member for ArF lithography according to any one of claims 1 to 5, wherein the thermal conductivity of the optical member at 22 ° C is 1.2 W / (m · ° C) or more.   The optical member for ArF lithography according to any one of claims 1 to 6, wherein the optical surface of the optical member has a Vickers hardness Hv of 6.5 GPa or more.   The optical member for ArF lithography according to any one of claims 1 to 7, wherein a virtual temperature Tf of the optical member is 900 to 1100 ° C.   An aluminum thin film is formed as a reflective film on the optical surface of the optical member for an ArF lithography mirror according to claim 1, and at least one dielectric thin film is formed as a protective film on the reflective film. A mirror for ArF lithography.   The mirror for ArF lithography according to claim 9, wherein the light reflectance at a wavelength of 193 nm is 80% or more. 11. The ArF lithography according to claim 10, wherein a decrease in light reflectance at a wavelength of 193 nm before and after the ArF laser light irradiation when irradiated with 10 ⁷ shot ArF laser light at 20 mJ / cm ² / pulse is 2% or less. mirror.