JP2001302255A - Method for manufacturing optical member, optical member, and projection exposure apparatus - Google Patents

Abstract

(57) [Summary] To improve the optical performance of an optical member. SOLUTION: A silicon compound is spouted from a central portion of a burner 1 and hydrolyzed in an oxyhydrogen flame to form a rotary target 2.
The ingot 7 is synthesized by melt vitrification above. At this time, the rotary target 2 is driven to swing in the X direction. The value obtained when the relative distance between the axis extending in the direction substantially parallel to the Z axis through the center of the burner 1 and the rotation center axis of the rotary target 2 becomes the largest due to the swing is D. When the radius of the ingot 7 is R, the swing stroke is determined so as to satisfy R / 5 ≦ D ≦ R.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for manufacturing an optical member, an optical member, and a projection exposure apparatus, and more particularly, to projecting a pattern formed on a mask onto a substrate using light having a wavelength of 400 nm or less. The present invention relates to a method for manufacturing an optical member made of quartz glass suitable for use in an illumination optical system or a projection optical system of a projection exposure apparatus, an optical member manufactured by the manufacturing method, and a projection exposure apparatus using the optical member.

[0002]

2. Description of the Related Art It is desired that an optical member used for an illumination optical system or a projection optical system of a projection exposure apparatus has a high transmittance with respect to light transmitted through the optical member. Because the optical system of the projection exposure apparatus is composed of a combination of a large number of optical members, the transmittance of each optical member must be reduced in order to suppress a decrease in the transmittance of the entire optical system. It is necessary to increase the

[0003] When an optical member having a low transmittance is used,
The heat generated by the optical member absorbing a part of the transmitted light increases. Then, the refractive index of the optical member becomes non-uniform or local thermal expansion of the optical member occurs. Any of these causes the performance of the optical system to deteriorate.

In addition, it is desired that the above-mentioned optical members have high refractive index homogeneity and low distortion. Reason,
This is because higher optical performance is required for the illumination optical system and the projection optical system in order to form a finer pattern image on the substrate. If the homogeneity of the refractive index of the optical member is low, the performance as designed cannot be achieved even if the polished surface accuracy of the optical member is high.

In view of the circumstances described above, the optical system of a high-resolution projection exposure apparatus that uses a light source that emits light having a wavelength of 400 nm or less for exposure is made of quartz glass, calcium fluoride, or the like. Is used. These materials have relatively high transmittance in the ultraviolet,
In addition to having relatively high refractive index homogeneity, it also has excellent properties of low distortion.

[0006] It has been found that when distortion is present in an optical member, light transmitted through the optical member undergoes birefringence, resulting in a decrease in optical performance. An optical member capable of obtaining higher optical performance by controlling the amount of birefringence is disclosed in Japanese Patent Application Laid-Open No. 8-107060 by the present applicant. In this publication, the birefringence amount (absolute value) is 2
nm / cm or less, that is, the amount of separation between ordinary light and extraordinary light when transmitted through an optical member having a thickness of 1 cm is set to 2 nm or less, and the distribution of birefringence along the radial direction of the optical member is set at the center. It is disclosed that high performance can be achieved by symmetrical configuration. Due to the use of the optical member and the shortening of the wavelength of the light source as described above, the pattern line width that can be formed by the projection exposure apparatus is becoming finer and faster.

[0007]

However, as the projection exposure apparatus is required to have a higher resolution, a desired optical performance may not be obtained even if an optical member having high optical characteristics as described above is used. .

An object of the present invention is to provide a method for manufacturing an optical member capable of obtaining higher optical performance, an optical member manufactured by the method, and a high resolution using the optical member. To provide a simple projection exposure apparatus.

[0009]

FIG. 1 shows an embodiment of the present invention.
The following invention will be described in association with. (1) According to the first aspect of the present invention, while rotating the rotary target 2 and the multi-tube burner 1 relative to each other, a silicon compound is ejected from the multi-tube burner 1 to hydrolyze in a oxyhydrogen flame, and the rotary target is rotated. The method is applied to a method for manufacturing an optical member 7 made of quartz glass to be melted and vitrified on the surface 2. Then, the value when the relative distance between the center of the multi-tube burner 1 and the central axis of the rotary target 2 becomes the largest along with the swing is D, and the radius of the manufactured optical member 7 is R. In some cases, the above-described object is achieved by performing swinging so as to satisfy the following expression. R / 5 ≦ D ≦ R (2) The invention according to claim 2 is applied to an optical member used in a specific wavelength band of 400 nm or less, and shows a distribution of a birefringence amount along a diameter direction of the optical member. The shape of the curve has an extreme value near the center of the optical member and has an extreme value in a region outside the effective use range of the optical member. (3) The invention according to claim 3 is applied to an optical member used in a specific wavelength band of 400 nm or less, and the rate of change in the amount of birefringence due to the diameter change of the optical member is 0.2 nm.
/ Cm / cm or less, the distribution of birefringence along the radial direction of the optical member is determined. (4) In connection with FIG. 2 showing an embodiment, the invention according to claim 4 is directed to a projection exposure apparatus that projects a pattern formed on a mask M onto a substrate W by a projection optical system PL. Applied. An optical member according to any one of claims 2 to 3 is used in the projection exposure apparatus.

The present inventors have found that the cause of the inability to sufficiently bring out the optical performance of the optical member is that the distribution of birefringence along the radial direction of the optical member has an extreme value. This `` having an extreme value '' means that a curve obtained by plotting the diameter of the optical member on the horizontal axis of the orthogonal coordinates and plotting the magnitude of the birefringence amount at each diameter on the vertical axis. This means that the curve has a local maximum and a local minimum.

When the distribution of the birefringence has an extreme value, the change of the birefringence in a portion near the extreme value becomes large. For example, when an optical member in which a distribution of a birefringence amount along a radial direction is small as an average value but a portion where a rate of change of the birefringence amount is locally large is used, the optical member has The amount of birefringence of the passing light changes relatively large depending on the transmitting position. Usually, the optical system is often constituted by a plurality of optical members. At this time, if there are a plurality of optical members whose birefringence amount locally changes as described above, the amount of birefringence increases each time light passes through each optical member, and the optical system as a whole optical system Performance tends to decrease.

Conventionally, when an optical member is evaluated based on the amount of birefringence, it is determined only by whether the amount of birefringence is large or small. However, in order to improve the performance of the optical system in which this optical member is used, a convex or concave locally present in the distribution curve of the birefringence along the radial direction of the optical member while reducing the amount of birefringence, That is, the present inventors have found that it is effective to eliminate the extreme value. In addition to the above, the rate of change (change rate) of the distribution curve of the amount of birefringence along the radial direction of the optical member is small, in other words, the gentle distribution curve draws up the performance of the optical system. Has been found to be effective.

To further explain the above,
Conventionally, there has been no concept of adding a sign to the amount of birefringence. Therefore, when measuring the amount of birefringence of an optical member, the amount of birefringence is measured at a plurality of points on the circumference of about 95% of the outer diameter of the optical member, and the maximum value of the measured values is the optical member for which Of the birefringence. However, when the distribution of the amount of birefringence of the optical member was measured, it had various distribution shapes, and it was not sufficient to manage the maximum value of the amount of birefringence at a specific diameter as described above. I understand. For example, in the case of quartz glass, a residual strain distribution occurs when the quartz glass is cooled due to a temperature distribution, an impurity distribution, or a SiO2 structure distribution at the time of synthesis, and this causes a distribution of birefringence. It is thought that it is. At this time, the form of birefringence differs depending on whether the strain is a compressive strain or a tensile strain.

The birefringence refers to a phenomenon in which one incident light is separated into an extraordinary ray and an ordinary ray when passing through an optically anisotropic body. The extraordinary ray indicates a ray having a different refractive index depending on the direction of the vibration surface when propagating through the optically anisotropic body. On the other hand, an ordinary ray indicates a ray having a constant refractive index regardless of the direction of the vibrating surface. There is a refractive index ellipsoid as a method of expressing that the refractive index changes depending on the direction of the vibrating surface of the light beam and the incident direction. Regardless of the direction of the vibrating surface and the incident direction of the light beam, if it has a constant refractive index,
The refractive index ellipsoid becomes a true sphere. Here, the direction in which the refractive index is small, that is, the minor axis direction of the refractive index ellipsoid is defined as the fast axis. Similarly, the direction in which the refractive index is high, that is, the major axis direction of the refractive index ellipsoid is defined as the slow axis. In the present invention,
Considering the direction of the fast axis when evaluating the distortion of the optical member,
If the direction of the fast axis is parallel to the radial direction of the optical member, it is expressed as plus, if it is perpendicular, it is expressed as negative, and the degree is expressed as a numerical value, and the distribution of strain along the radial direction of the optical member is expressed. It was measured. At this time, if the measured value of birefringence is small, the fast axis may not always be completely parallel or perpendicular to the diameter but may have an inclination. In this case, a sign closer to parallel than an angle of 45 degrees with respect to the diametrical direction may be treated with + sign, and a sign close to perpendicular with-may be treated. When evaluating the distribution of the amount of birefringence in the optical member, the directionality (plus or minus) and the magnitude of the strain change gradient along the radial direction of the optical member are very important indicators.

In the section of the means for solving the above-mentioned problems, which explains the configuration of the present invention, the drawings of the embodiments of the present invention are used for easy understanding of the present invention. However, the present invention is not limited to this.

[0016]

FIG. 1 schematically shows a structure of a synthesis furnace to which a method for manufacturing an optical member according to an embodiment of the present invention is applied. Synthesized. In the following description, the X axis is taken in the horizontal direction of the paper surface of FIG. 1, the Y axis is taken in a direction perpendicular to the paper surface, and the Z axis is taken in the up and down direction, and the directions along these X, Y and Z axes are X Direction, Y direction, and Z direction. Also, for example, Z
The direction in which the coordinate value is reduced in the direction (a downward direction along the plane of FIG. 1) is referred to as a “minus Z direction”.

The burner 1 has a multi-tube structure (a structure in which a plurality of tubes are bundled into one), and is arranged so that the tip thereof is directed to a rotating target 2 disposed below the tube. The furnace wall is composed of a furnace frame 4 and a refractory 3. The furnace wall has an observation window (not shown) for visually observing the inside of the furnace, and a monitor for monitoring the inside of the furnace with an infrared camera 9. A window 5 and an exhaust system 6 are provided. The burner 1 is moved in the X direction and Y direction in FIG.
It is movably supported in at least one of the directions.

At the lower part of the furnace, a rotary target 2 for forming an ingot 7 is rotatably supported by a support shaft 8. A motor 13 is connected to the support shaft 8 so that the rotary target 2 can be rotated at an arbitrary rotation speed. The rotating target 2 can be swung in at least one of the X direction and the Y direction in FIG. 1 by the driving device 10. That is, the rotary target 2 is set to X in FIG.
By swinging in only one of the direction and the Y direction, swinging in a one-dimensional direction along a plane orthogonal to the rotation axis of the rotary target 2 becomes possible. Further, by swinging the rotary target 2 in the X direction and the Y direction with a combination of a predetermined swing stroke, swing cycle, and swing phase, swing in the two-dimensional direction becomes possible. Drive device 10
Is composed of a motor 11, a feed screw 12, and the like.

The rotary target 2 and the driving device 10
It is fixed to a fixed column 15 via a Z drive device 14.
The rotary target 2 and the driving device 10 are the Z driving device 1
4 is driven in the ± Z direction in FIG.

An oxygen-containing gas and a hydrogen-containing gas are ejected from the burner 1, mixed, ignited, and an oxyhydrogen flame is formed. From the center of the burner 1, a silicon compound as a raw material is diluted with a carrier gas and ejected. Then
The above raw material is hydrolyzed in an oxyhydrogen flame to generate fine silica glass particles (soot).

The above-mentioned soot is deposited on the oscillating and rotating rotary target 2. The soot deposited on the rotating target 2 is heated and melted and vitrified by the oxyhydrogen flame. By performing the above process for a long time, the transparent quartz ingot 7 is formed on the rotating target 2. At this time, the upper part of the ingot 8 is covered with the flame emitted from the burner 1. The distance between the burner 1 and the glass composite surface of the ingot 7 (top of the ingot) is always kept constant. That is, as the ingot 7 grows, the rotating target 2 is lowered in the minus Z direction.

The raw material to be ejected from the burner 1 is S
SiCl such as iCl4, SiHCl3, SiF
4, silicon fluoride such as Si2F6, hexamethyldisiloxane, octalmethylcyclotetrasiloxane,
Siloxanes such as tetramethylcyclotetrasiloxane, organosilicon such as silanes such as methyltrimethoxysilane, tetraethoxysilane and tetramethoxysilane, and silicon compounds such as SiH4 and Si2H6 can be used.

Since transparent quartz glass has high viscosity even at high temperatures, it does not flow and drip without using a container or the like. That is, the transparent quartz glass grows while maintaining its own shape.

Incidentally, the flame caused by the oxyhydrogen gas ejected from the burner 1 has a temperature distribution along the radial direction of the ingot 7. The reason is that the above-mentioned raw material is ejected from the central portion of the burner 1, while the oxyhydrogen gas is ejected from the periphery thereof to emit a flame. That is, while the temperature of the oxyhydrogen flame is 2,000 ° C. or higher, the temperature is only about several hundred degrees Celsius in the vicinity of the raw material ejection part even if the raw material is performing a combustion reaction. Also, the outer part of the flame,
That is, the temperature of the portion where no flame exists is naturally lower than the temperature of the flame. Therefore, the temperature distribution along the radial direction of the ingot 7 has a distribution such that the temperature distribution is low near the center, becomes higher toward the outer diameter side, and sharply decreases further toward the outer diameter side.

The state of distribution of the amount of birefringence in the ingot is as follows:
Similar to the refractive index distribution, it is determined depending on the state of inhomogeneity of the glass structure due to the existence of the above-mentioned temperature distribution when the ingot 7 is synthesized, the distribution of impurities such as OH groups or chlorine, and the like. That is, residual stress exists in the ingot 7 that has been cooled due to the presence of the temperature distribution and the impurity distribution as described above. This residual stress has a distribution similar to the temperature distribution described above, and has an extreme value in a cloth curve showing a distribution of the residual stress along the radial direction from the center of the ingot 7 to the outer periphery. It is considered that the reason why the distribution of the residual stress exists is as follows. That is, in the process of synthesizing the ingot 7, the temperature of the central portion of the ingot 7 is lower than the temperature of the portion immediately outside thereof, so that the density of the central portion becomes relatively high and the portion immediately outside thereof is rough. . Further, since the temperature of the outer portion is lower, the density increases again. In accordance with the existence of such a residual stress distribution, a residual strain distribution exists in the ingot 7. Then, a distribution of the amount of birefringence is generated according to the existence of the distribution of the residual strain. Having an extreme value in the distribution of the residual strain means that there is a change in the directionality of the stress (when the differential coefficient is obtained, the sign is inverted before and after the extreme value), and the amount of change in strain is It leads to being big. That is, when the distribution of the birefringence along the radial direction of the ingot 7 is viewed, the fluctuation of the birefringence increases.

The present inventors have found that it is effective to improve the performance of the optical member by setting the above-mentioned portion where the amount of birefringence varies greatly outside the effective use range of the optical member. .

The extreme value of the distribution curve indicating the distribution of the residual strain along the radial direction of the ingot 7 occurs near the boundary between the part where the raw material is ejected and the part where the flame exists. In order to drive the position of the extreme value out of the effective use range of the optical member, in the synthesis furnace shown in FIG. 1, the rotary target 2 is swung in the ± X direction of FIG. Move. At this time, the swinging operation of the rotary target 2 may be made to swing with the center position of the synthesis furnace as the swing center, or may be made to swing with the center position of the synthesis furnace as the swing end. In this way, the effect of the low temperature portion at the center of the burner is equalized on the upper surface (glass composite surface) of the ingot 7, whereby the amount of variation in the distribution of residual strain can be reduced. Can be moved out of the effective use range of the optical member. At this time, since the distribution width of the temperature on the glass composite surface is also narrowed, the absolute amount of the residual strain in the ingot 7 can be reduced.

If the diameter of the ingot 7 to be synthesized is increased and the diameter of the hollow is reduced, the influence of the temperature distribution is relatively reduced, and the variation of the residual strain existing along the radial direction is made substantially flat. be able to. However, even if the amount of fluctuation of the residual strain is small, if the difference between the maximum value and the minimum value of the residual distortion in the optical member is large, the performance of the optical system assembled using this optical member is reduced. The desired optical performance cannot be obtained. Specifically, the difference between the maximum value and the minimum value of the birefringence amount distributed in the optical member is 2 nm /
cm or less.

Here, a phase modulation method, which is a method for measuring the amount of birefringence performed in place of the measurement of the residual strain in the embodiment of the present invention, will be described. A measurement system used in the measurement by the phase modulation method includes a light source, a polarizer, a phase modulation element, an analyzer, and a photoelectric conversion element. As a light source, a He-Ne laser, a laser diode, or the like can be used. As a phase modulation element, a photoelastic modulator (hereinafter, referred to as a phase modulation element)
In this specification, a photoelastic modulator is referred to as “PEM”. PEM is a quartz plate, CaF2, or Z
Vibration can be generated by applying an AC voltage of several tens of kHz. By changing the frequency of the applied AC voltage, the phase difference between the ordinary ray and the extraordinary ray separated by birefringence when light passes through this PEM is continuously changed from 0 radians to 2π radians. be able to. In other words, continuously changing the polarization state of the light obtained by recombining the ordinary and extraordinary rays separated at the time of transmission through the PEM from linearly polarized light to circularly polarized light, and from circularly polarized light to linearly polarized light. Can be.

Light emitted from the light source passes through the polarizer and becomes linearly polarized light. This linearly polarized light is the same as the PEM described above.
Changes the polarization state. The sample to be measured is arranged between the phase modulation element and the analyzer. When measuring, the signal peak is detected by detecting the signal output from the photoelectric converter while rotating the sample around the light beam incident on the measurement target point on the sample, and the amplitude at that time is measured. The direction of the fast axis or the slow axis and the birefringence phase difference (the phase difference between the ordinary ray and the extraordinary ray) are determined.
The details of the method of measuring the amount of birefringence by the above-described phase modulation method are described in detail in the following literature (Etsuhiro Mochida: “Birefringence Measurement and Application by Phase Modulation Method”, Optical Technology Contact Vol. 27, No. .3,198
9).

If a Zeeman laser is used as a light source in the above measurement system, it is possible to perform measurement without rotating the sample. As other measurement methods using a Zeeman laser, a phase shift method and an optical heterodyne interferometry can also be used.

In addition, although the measurement accuracy is slightly inferior, it is also possible to perform measurement by several methods as described below.
The measuring device used in the rotating analyzer method has a configuration in which a sample between a light source and a photoelectric conversion element is sandwiched between a polarizer and a rotating analyzer. A signal output from the photoelectric conversion element is measured while rotating the analyzer disposed after the sample to be measured, and a phase difference can be obtained from the maximum value and the minimum value of the signal output from the photoelectric conversion device.

In the phase compensation method, a light source, a polarizer, a sample, a phase compensator, an analyzer, and a photoelectric conversion element are arranged. The polarizer and the analyzer are arranged such that their axes are orthogonal to each other. Since the linearly polarized light incident on the sample to be measured becomes elliptically polarized light due to the birefringence of the sample, the elliptically polarized light is returned to linearly polarized light by adjusting the phase compensator. By adjusting the phase compensator, the signal output from the photoelectric conversion element becomes almost zero. The phase compensation value at the time of best extinction corresponds to the amount of birefringence.

In addition to the above measurement method, a simple method of placing a standard sample in a crossed Nicols optical system and comparing the standard sample with the sample to be measured is also possible if the thickness of the sample to be measured is sufficient.

The measured values of the amount of birefringence obtained by the above-mentioned measuring method are plus when the fast axis is in a direction parallel to the diameter of the optical member, and when the fast axis is in a direction perpendicular to the optical member. Is a minus sign. If the measured value of birefringence is small, the direction of the fast axis is not necessarily parallel to the diameter of the optical member,
Alternatively, there is a case where the camera does not face a vertical direction and has an inclination.
In such a case, those having an inclination of 45 degrees or less (close to parallel) with respect to the direction along the diameter of the optical member are plus signs, and those having an inclination exceeding 45 degrees (close to vertical) are minus signs. It should be handled with the addition of.

Next, an example of a specific configuration of the projection exposure apparatus according to the present invention will be described. FIG. 2 schematically shows the configuration of the projection exposure apparatus according to the present invention. FIG. 2 shows a state in which a light beam emitted from the light source 100 is developed on a straight line without illustration of a reflection optical system or the like appropriately arranged in an optical path until the light beam enters the photosensitive substrate W. .
FIG. 2A is a diagram showing a configuration when the projection apparatus is viewed from directly above, and FIG. 2B is a diagram showing a configuration when the projection exposure apparatus of FIG. 2A is viewed from the side. FIG. As shown in FIG. 2, from a light source 100 including an excimer laser oscillator, a solid-state laser oscillator, and the like, 248 nm (KrF excimer laser), 193 nm (ArF excimer laser),
A substantially parallel light beam having a wavelength such as 157 nm (F2 laser) is emitted, and the cross-sectional shape of the parallel light beam at this time is rectangular. The parallel light beam emitted from the light source 100 enters a beam shaping optical system 20 that shapes the light beam into a light beam having a predetermined cross-sectional shape.

The beam shaping optical system 20 is composed of two cylindrical lenses 20A and 20B having a refractive index in a direction perpendicular to the plane of FIG. 2A, that is, in a direction along the plane of FIG. 2B. . The cylindrical lens 20A on the light source side has a negative refractive index, and has a function of diverging a light beam in a direction along the plane of FIG. 2B. The cylindrical lens 20B on the irradiated surface side has a positive refractive index,
The divergent light beam incident from the cylindrical lens 20A on the light source side is condensed and converted into a parallel light beam. The light beam transmitted through the beam shaping optical system 20 in this manner has its light beam width expanded in the direction along FIG. 2B, and the beam shape is shaped into a rectangular shape. The configuration of the beam shaping optical system 20 may be a combination of a plurality of cylindrical lenses having a positive refractive index other than those described above, or an anamorphic optical system using a prism or the like. May be configured.

The light beam shaped by the beam shaping optical system 20 enters the first relay optical system 21. The first relay optical system 21 has a front group having a positive refractive power composed of two convex lenses 21A and 21B, and a rear group having a positive refractive power composed of two convex lenses 21C and 21D. The front group of the first relay optical system 21 forms a light source image I at a focal position on the mask M side (rear side) of the front group. The rear group of the first relay optical system 21 is arranged such that its front focal point coincides with the rear focal position of the front group. The first relay optical system configured as described above has a function to conjugate the exit surface of the light source 100 and the entrance surface of the optical integrator 30 described later. By disposing the first relay optical system 21 between the beam shaping optical system 20 and the optical integrator 30, the deviation of the light beam illuminating the optical integrator 30 due to the angular deviation of the light from the light source 100 is corrected. The tolerance to the angular deviation of the light from the light source is increased.

The light beam transmitted through the first relay optical system 21 is
The light is incident on an optical integrator 30 for forming a plurality of light source images. The optical integrator 30 is configured by arranging a plurality of biconvex lens elements having a substantially square light transmitting surface shape,
The optical integrator 30 as a whole has a rectangular light transmitting surface shape. Each lens element has the same curvature (refractive index) in the direction along the plane of FIG. 2A and the direction along the plane of FIG. 2B.

Therefore, the optical integrator 3
The parallel luminous flux transmitted through the individual lens elements constituting 0 is
The light is condensed to form a light source image on the emission side (rear side) of each lens element. In this way, a plurality of light source images corresponding to the number of lens elements are formed at the emission side position A1 of the optical integrator 30.
A secondary light source is formed.

As described above, the light beams from the plurality of secondary light sources formed by the optical integrator 30 are:
The light is condensed by the second relay optical system 40 and further enters an optical integrator 50 that forms a plurality of light source images. The optical integrator 50 is configured by arranging a plurality of biconvex lenses having a rectangular light transmitting surface shape. The light transmitting surface shape of each lens element is configured to be similar to the light transmitting surface shape of the optical integrator 30. The light transmitting surface of the entire optical integrator 50 has a square shape. Each lens element constituting the optical integrator 50 is shown in FIG.
The direction along the plane of FIG. 2A and the direction along the plane of FIG. 2B have the same curvature (refractive index).

As described above, the light emitted from the optical integrator 30 is condensed by the individual lens elements constituting the optical integrator 50, and a light source image is formed on the exit side (rear side) of each lens element. . Therefore, a plurality of light source images arranged in a square shape are formed at the emission side position A2 of the optical integrator 50, and a tertiary light source is substantially formed here.

The second optical system 40 conjugates the incident surface position B1 of the optical integrator 30 with the incident surface position B2 of the optical integrator 50, and sets the exit surface position A1 of the optical integrator 30.
And the exit surface position A2 of the optical integrator 50 are conjugated.

An aperture stop A having a predetermined aperture shape is located at or near the position A2 where the tertiary light source is formed.
S is provided, and a light beam from a tertiary light source formed in a circular shape by the aperture stop AS is condensed by a condenser optical system 60 which is a condensing optical system. This light beam is formed into an elongated rectangular beam shape by a slit-shaped opening (not shown) called a reticle blind, and the mask M
Is illuminated in a thin strip.

The mask M is held on a mask stage MS, and the photosensitive substrate W is held on a substrate stage WS. The mask M and the photosensitive substrate W are provided with a projection optical system P
A reduced image of the pattern of the mask M, which is arranged conjugate with respect to L and is illuminated in a thin strip, is projected onto the photosensitive substrate W by the projection optical system PL.

In the actual projection exposure operation by the projection exposure apparatus configured as described above, the mask stage MS
The substrate stage WS and the substrate stage WS move in opposite directions as indicated by arrows in FIG. 2B at a speed ratio equal to the imaging magnification of the projection optical system PL. Thereby, a reduced image of the mask M is formed on the photosensitive substrate W.

The projection optical system PL has, for example, the configuration shown in FIG. Hereinafter, the projection optical system PL will be described with reference to FIG.
Will be described. The projection optical system PL includes a mask M
, The first lens group G1 having a positive power, the second lens group G2 having a positive power, the third lens group G3 having a negative power, the fourth lens group G4 having a positive power, and The fifth lens group G5 having a positive power and the sixth lens group G6 having a positive power
And The stop AS1 is arranged on the incident surface side of the concave lens L54 in the fifth lens group G5. The projection optical system PL is almost telecentric on the mask M side and the photosensitive substrate W side and has a reduction magnification.

In the projection optical system PL described above, L45, L46, L63, L6
5, L66 and L67 are made from a single crystal of calcium fluoride. Other lenses are made from quartz glass.

According to the method for manufacturing an optical member described above,
An optical member constituting the projection optical system PL is manufactured. At this time, as described above, the distribution of the amount of birefringence is measured for each optical member, and a member suitable for obtaining desired optical performance is selected based on the data of the amount of birefringence. As an optical member constituting the projection optical system PL of the above-described projection exposure apparatus,
The change gradient of the birefringence distribution curve along the radial direction is relatively small, and the birefringence difference (the maximum and minimum values of the birefringence distributed along the radial direction of the optical member, Is less than 2 nm / cm.

As described above, by reducing the amount of birefringence of each of the optical members constituting the projection optical system PL and the amount of fluctuation of the amount of birefringence along the radial direction, the entire projection optical system PL The variation in the amount of birefringence is reduced, and the absolute amount is also reduced. Accordingly, the wavefront aberration of the entire projection optical system PL is reduced to a minimum, and a projection exposure apparatus with high resolution can be obtained.

[0051]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to embodiments.

Example 1 A synthetic quartz glass was manufactured by a direct method using the synthesis furnace shown in FIG. Silicon tetrachloride was used as a raw material to be ejected from the center of the burner 1. Then, a quartz glass ingot having an outer diameter of 400 mm was synthesized while swinging the rotary target 2 in the X direction. The burner 1 was fixed near the center of the synthesis furnace, and the rotary target 2 was oscillated by swinging around the center of the synthesis furnace. In synthesizing one quartz glass ingot, the swing stroke was changed stepwise over time to 50 mm, 80 mm, and 150 mm.

After the synthesis of the quartz glass ingot was completed, the ingot was allowed to cool in the air, and cylindrical test pieces 1, 2, and 3 having an outer diameter of 400 mm and a thickness of 100 mm were cut out. At this time, the test piece 1 was cut from a portion synthesized with a swing stroke of 50 mm, the test piece 2 was cut from a portion synthesized with a swing stroke of 80 mm, and the test piece 3 was cut from a portion synthesized with a swing stroke of 150 mm.

The test pieces 1 to 3 cut out as described above were heated to 1,000 ° C.
After holding at ゜ C for 10 hours, 5 at a rate of 10 ゜ C / hour
Annealing was performed under the conditions of gradually cooling to 00 ° C. and then cooling in the air. The distribution of the refractive index along the radial direction of each of the annealed test pieces 1 to 3 was measured by the phase modulation method.

FIG. 4A shows the results of the above-mentioned measurement. The maximum values of the birefringence were 2 nm / cm, 1.6 nm / cm and 1 nm / c for the test pieces 1, 2, and 3, respectively.
m. From these measurement data, the rate of change of the birefringence amount along the radial direction of each of the test pieces 1, 2, and 3 was calculated for each 1 cm. As a result, the maximum values were 0.40 nm / cm / cm and 0. 20 nm / cm / c
m, 0.05 nm / cm / cm.

The distribution curves of the birefringence shown in FIG. 4A will be further described. These distribution curves show the distribution along the radial direction of the optical member, and the distribution along the diameter direction is shown in FIG. ) Has a line-symmetric shape with the vertical coordinate axis of the graph shown in FIG. When considering the entire distribution along the diameter direction of the optical member, the birefringence distribution curve of the test piece 1 has a concave portion near the center and a convex portion at an intermediate diameter portion near a radius of 80 mm. That is, the distribution curve of the amount of birefringence has an inflection point (extreme value) near the center of the optical member and the intermediate diameter portion. Similarly, the birefringence distribution curve of the test piece 2 has an extreme value near the center of the optical member and an intermediate diameter portion near a radius of 120 mm. However, test piece 1
, The birefringence amount difference (difference between the maximum value and the minimum value of the birefringence amount) is small. The birefringence distribution curve of the test piece 3 has an extreme value only near the center of the optical member. The rate of change of the amount of birefringence is small, and the distribution curve is gentle as a whole.

Subsequently, the operation of synthesizing the ingot and cutting out test pieces having an outer diameter of 400 mm and a thickness of 100 mm by the same synthesizing method (oscillating stroke) as the above-described synthesizing method of the test pieces 1 to 3 was repeated, and each operation was repeated 50 times. Pieces, total 1
Fifty test pieces were made. After annealing all of these test pieces under the same conditions as described above, the distribution of birefringence was measured. Then, based on the obtained measurement data, an index for predicting the optical performance of the entire projection optical system PL when the projection optical system PL was manufactured using these test pieces was calculated as follows. .

In order to calculate the index of the optical performance, a Strehl value calculation formula shown by the following formula (1) was used.

(Equation 1)

A specific example of the process of calculating the Strehl value will be described. In the 150 test pieces mentioned above,
Each is given a unique code. For example, 50 test pieces obtained by the same synthesis method as that of the test piece 1 include 1-1, 1-2,...
A reference numeral of -50 is attached. Regarding those obtained by the same synthesis method as that of the test piece 2, 2-1 and
2-2,..., 2-49, 2-50.
, 3-2,..., 3-49, 3-
The reference numeral 50 is added. Hereinafter, in the present example, 50 pieces obtained by the same synthesis method as that of the test piece 1 were obtained.
Test pieces as "first sample", test piece 2
The 50 test pieces obtained by the same synthesis method as the synthesis method of “2nd sample” and the 50 test pieces obtained by the synthesis method similar to the synthesis method of the test piece 3 are the “third sample”. Respectively.

The distribution of the amount of birefringence was measured in advance for all of the first sample, the second sample, and the third sample, and each was evaluated by the method described below.

Incidentally, the projection optical system PL shown in FIG.
Is composed of 29 lenses. Among them, 23 sheets were made of quartz glass, and the remaining 6 sheets (L45, L4
6, L63, L65, L66, L67) are manufactured from a single crystal of calcium fluoride as described above. When the first sample is evaluated, for example, “the lens L11 is the test piece 1-2, and the lens L12 is the test piece 1-, while observing the distribution of the amount of birefringence previously measured and the effective diameter (outer diameter) of each lens. At 50, the lens L13 is connected to the test piece 1-
1, etc. "and 29 lenses are associated with each sample. The reason for determining the association in this way is that the distribution of the birefringence varies from test piece to test piece,
This is because it is important to determine which test piece to use according to the lens diameter, the power, and the like, in order to enhance the performance of the entire projection optical system PL.

Based on the correspondence determined as described above, the measurement result of the distribution of the amount of birefringence corresponding to each of the lenses L11 to L610 is input to a computer for calculating a Strehl value. Regarding the distribution of the birefringence of the six lenses manufactured from a single crystal of calcium fluoride, theoretical values may be input to a computer, or the measurement of a sample having a standard distribution of birefringence may be performed. You may enter a value.
Alternatively, samples for these six lenses may be prepared at the same time and evaluated together with a sample made of quartz glass.

Then, a ray passing point on each lens is obtained by a method similar to ray tracing used for calculating aberrations of the optical system, for example, on the optical axis, paraxial region, off-axis, etc., and corresponds to the point. The Strehl value is determined by substituting the amount of birefringence into equation (1). That is, the projection optical system PL at various incident angles
A plurality of Strehl values corresponding to a plurality of light rays incident on the sample are determined, and the lowest value among them is determined as the Strehl value of the combination of the sample. This was performed for each of the first sample, the second sample, and the third sample.

As a result of preparing the samples described above, measuring the birefringence distribution of each sample, selecting a sample corresponding to each lens, and calculating the Strehl value, 0.85 for the first sample and 0 for the second sample. .95, and 0.99 for the third sample. In general, it is known that when the Strehl value is 0.95 or more, the projection optical system can obtain desired performance. Therefore, the performance of the projection optical system can be made close to the design value by synthesizing the ingot and manufacturing the optical member by the same manufacturing method as the manufacturing method of the second sample and the third sample.

Of the above samples, the test piece selected from the third sample having the best Strehl value was actually polished to produce a lens, and the projection optical system PL shown in FIG. 3 was assembled and evaluated. Was. As a result, a projection exposure apparatus using the projection optical system PL could obtain a resolution of 0.18 μm.

Referring now to FIG. 1, the relative distance between the center axis of the burner 1 extending in a direction substantially parallel to the Z axis and the center axis of rotation of the rotary target 2 will be described. As described above, the rotating target 2 is swingably driven by the driving device 10 when the ingot 7 is combined. This relative distance becomes maximum when the rotating target 2 reaches the swinging end, and the maximum value of the relative distance in the process of synthesizing the ingot 7 is 25 mm at the time of synthesizing the first sample ingot, and the maximum value of the second sample ingot. Sometimes 40m
m, 75 mm when the third sample ingot is synthesized. When the ratio between the maximum value of these relative distances and the radius R (= 200 mm) of the ingot 7 is obtained, R / 8 at the time of combining the ingot for the first sample, R / 5 at the time of combining the ingot for the second sample, and R / 3 at the time of combining the ingot for the second sample. R × (3/8) at the time of ingot synthesis. That is, in this embodiment, the maximum value of the relative distance is set to R / 5 or more, the ingot 7 is synthesized, and a lens is manufactured from the ingot 7, so that the optical performance of the projection optical system PL is almost as designed. And could be. At this time, looking at the distribution of birefringence along the diameter direction of the optical member obtained by cutting out the ingot 7, the rate of change of the distribution curve is small,
It was confirmed that it was 0.2 nm / cm / cm or less.
In addition, as shown in the graph of FIG. 4A, the shape of the distribution curve of the amount of birefringence along the diameter direction of the optical member has an extreme value only at the center of the optical member, as in the test piece 3. It has been confirmed that the optical performance can be further improved by having. Note that the maximum value of the relative distance is desirably smaller than R, because otherwise, the quartz glass fine particles generated at the tip of the burner 1 will come off the rotary target 2.

By the way, it is difficult to completely eliminate variations in manufacturing conditions and the like, and the distribution curve of the amount of birefringence along the diametrical direction of the optical member shows a certain tendency as a whole, but the cut ingot or cut position. Some variation occurs due to the above. For this reason, when looking at the distribution of the birefringence, some of the above criteria (the birefringence difference is 2 nm /
cm or less, and the rate of change of the distribution curve is 0.2 nm / cm / cm.
The following may be deviated. In such a case, the optical member may satisfy the above criteria by producing a lens having a small effective diameter (outer diameter). That is, the projection optical system PL shown in FIG.
There are various diameters of the lens constituting the lens, for example, a lens having the smallest diameter is about 100 mm, and a lens having the largest diameter is about 300 mm.

When a lens having a small diameter is manufactured, only the central portion of the optical member is used. For example, when manufacturing a lens having a diameter of about 100 mm, the above criterion may be applied only to a range of about 100 mm in the optical member. Conversely, a lens having a large outer diameter may be manufactured from an optical member in which the variation width of the birefringence amount is narrow and good homogeneity is obtained over a wide range. As described above, by determining the use of the optical member according to the distribution of the amount of birefringence, the yield can be increased. For example, FIG.
Even if the optical member has a birefringence distribution similar to that of the test piece 1 in the middle and has an extreme value at the intermediate diameter portion of the optical member, the diameter of the lens manufactured from this optical member (effective If the usage range is small, it may be usable.
That is, by making the extreme value exist in a region outside the effective use range, it is possible to suppress a decrease in the optical performance of the lens and to reduce the possibility of wasting the optical member.

Example 2 A synthetic quartz glass was manufactured by a direct method using the synthesis furnace shown in FIG. Silicon tetrachloride was used as a raw material to be ejected from the center of the burner 1. Then, a quartz glass ingot having an outer diameter of 400 mm was synthesized while swinging the rotary target 2 in the X direction. At this time, in the first half of the synthesis process of the quartz glass ingot, the burner 1 was fixed near the center of the synthesis furnace, and in the second half, the burner was moved by 15 mm in the plus X direction from near the center of the synthesis furnace and fixed. In addition, the rotary target 2 was swing-driven with a swing stroke of 50 mm with the vicinity of the center of the synthesis furnace as the swing center throughout the whole synthesis process of the quartz glass ingot.

The cooling conditions after the completion of the quartz glass ingot were the same as in Example 1, and were allowed to cool in the air. Thereafter, a test piece 1 having an outer diameter of 400 mm and a thickness of 100 mm from a portion synthesized in the first half of the quartz glass ingot synthesis process,
Test pieces 2 of the same size were cut out from the part synthesized in the latter half. Thereafter, annealing was performed under the same conditions as in Example 1. Then, the distribution of the refractive index along the radial direction of each of the annealed test pieces 1 and 2 was measured by a phase modulation method.

FIG. 4 (b) shows the above measurement results. The maximum (absolute value) of the amount of birefringence was 1.6 nm / cm and 1.0 nm / cm for the test pieces 1 and 2, respectively.
It has become. Further, based on these measurement data, the rate of change of the birefringence amount along the radial direction of each of the test pieces 1 and 2 was calculated for each 1 cm.
It was 40 nm / cm / cm and 0.20 nm / cm / cm.

The birefringence distribution curve shown in FIG. 4B also shows the distribution along the radial direction of the optical member as described in the first embodiment, and the distribution along the diameter direction is shown in FIG. It has a line-symmetric shape with the vertical coordinate axis of the graph shown in FIG. Considering the entire distribution along the diameter direction of the optical member, the birefringence distribution curve of the test piece 1 has a concave portion near the center and a convex portion at an intermediate diameter portion near a radius of 130 mm. That is, the distribution curve of the amount of birefringence has an inflection point (extreme value) near the center of the optical member and at the intermediate diameter portion.
have. The birefringence distribution curve of the test piece 2 is
It has an extreme value only near the center of the optical member. And
The rate of change of the amount of birefringence is small, and the distribution curve is gentle as a whole.

Subsequently, the test pieces 1 and 2 described above
The operation of synthesizing an ingot and cutting out test pieces having an outer diameter of 400 mm and a thickness of 100 mm was repeated by the same synthesizing method as in the above method, whereby 50 test pieces each, a total of 100 test pieces, were prepared. That is, in the process of synthesizing the ingot for the test piece 1, the burner 1 is fixed near the center of the synthesis furnace, and the rotary target 2 is swung with a swing stroke of 50 mm around the center of the synthesis furnace as the swing center. Dynamically driven. Then, in the process of synthesizing the ingot for the test piece 2, the burner is moved by 15 mm in the plus X direction from the vicinity of the center of the synthesis furnace, and is fixed.
Was swing-driven with a swing stroke of 50 mm with the vicinity of the center of the synthesis furnace as the swing center. Hereinafter, in the present example, 50 test pieces created by the same method as the method for synthesizing test piece 1 are referred to as “first samples”, and 50 test pieces created by the same method as the method for synthesizing test piece 2 are used. Is referred to as a “second sample”.

After annealing all of the first and second samples under the same conditions as described in Example 1, the distribution of the amount of birefringence was measured. Then, based on the obtained measurement data, for predicting the optical performance of the entire projection optical system PL when the projection optical system PL is manufactured using the test pieces in the first sample and the second sample. The index, that is, the Strehl value described in Example 1 was calculated. A specific example of the Strehl value calculation process is the same as that described in the first embodiment, and a description thereof will not be repeated.

As a result of preparing the above-mentioned samples, measuring the birefringence distribution of each sample, selecting a sample corresponding to each lens, and calculating the Strehl value, 0.85 for the first sample and 0 for the second sample. It became .95. Therefore, it is expected that the performance of the projection optical system can be made close to the design value by synthesizing the ingot and manufacturing the optical member by the same manufacturing method as the manufacturing method of the second sample.

Of the above samples, a test piece selected from the second sample having a good Strehl value was actually polished to produce a lens, and a projection optical system PL shown in FIG. 3 was assembled and evaluated. . As a result, a projection exposure apparatus using the projection optical system PL could obtain a resolution of 0.19 μm.

Referring now to FIG. 1, the relative distance between the center axis of the burner 1 and the center axis of the rotary target 2 in the second embodiment will be described. As described above, the rotating target 2 is swingably driven by the driving device 10 when the ingot 7 is combined. This relative distance becomes the largest when the rotating target 2 reaches the swing end, and the maximum value of the relative distance in the process of synthesizing the ingot 7 is 25 m at the time of producing the first sample ingot.
m, 40 mm during the synthesis of the second sample ingot. The maximum value of these relative distances and the radius R of the ingot 7
(= 200 mm), it is R / 8 at the time of synthesizing the first sample ingot, and R / 5 at the time of synthesizing the second sample ingot.

That is, in the present embodiment, the maximum value of the relative distance is set to R / 5 or more, the ingot 7 is synthesized, and a lens is manufactured from the ingot 7, thereby substantially reducing the optical performance of the projection optical system PL. It was able to be as designed. At this time, the distribution of the birefringence amount along the diameter direction of the optical member obtained by cutting out the ingot 7 showed that the change in the distribution curve was small and was 0.2 nm / cm / cm or less. Also, as shown in the graph of FIG. 4B, the shape of the distribution curve of the birefringence amount along the diameter direction of the optical member has an extreme value only at the central portion of the optical member as in the test piece 2. It has been confirmed that the optical performance of the projection optical system PL is improved by having the optical element.

In the second embodiment, as described at the end of the first embodiment, a predetermined standard (birefringence difference of 2 nm / cm or less, birefringence distribution (The rate of change of the curve is 0.2 nm / cm / cm or less)
It can be used even if it does not satisfy. That is, it is only necessary that the above-mentioned criteria be satisfied within the effective use range of the lens manufactured from this optical member.

Further, even in the case of an optical member having a birefringence distribution having an extreme value at an intermediate diameter portion of the optical member, similar to the test piece 1 in FIG. If the diameter (effective use range) of the lens manufactured from the member is small, it may be usable in the same manner as described in the first embodiment. That is, by making the extreme value exist in a region outside the effective use range, a decrease in the optical performance of the lens can be suppressed, and the probability of wasting the optical member is reduced.

In the above embodiment, an example in which the target 2 is swung in the X-axis direction in the process of synthesizing the ingot 7 shown in FIG. 1 has been described. It may be swung two-dimensionally on the XY plane. Alternatively, the burner 1 may be swung in a one-dimensional direction or a two-dimensional direction, or both may be driven to relatively swing.

Instead of rotating the rotary target 2, the burner and the target may be relatively rocked in a two-dimensional direction along the XY plane.

The optical member manufactured by the method of manufacturing an optical member according to the present invention can be used not only as a lens constituting a projection optical system of a projection exposure apparatus but also as a material for a lens constituting an illumination system. . Further, it can be used as a material for other optical elements such as a prism as well as a lens.

In the correspondence between the above-described embodiments and the claims, the burner 1 constitutes a multi-tube burner.

[0085]

As described above, according to the present invention, the following effects can be obtained. (1) According to the first aspect of the present invention, D is a value when the relative distance between the central axis of the multi-tube burner and the central axis of the rotating target becomes the largest along with the swing, When the radius of the manufactured optical member is R, R /
By swinging so as to satisfy 5 ≦ D ≦ R, it is possible to reduce the variation width of the distribution of the birefringence amount along the diameter direction of the optical member and reduce the rate of change of the distribution of the birefringence amount. The optical performance can be improved. (2) According to the invention described in claim 2, the shape of the distribution curve indicating the distribution of the birefringence amount along the diameter direction of the optical member is as follows:
By having an extreme value near the center of the optical member, the rate of change in the distribution of birefringence along the diameter direction of the optical member can be reduced. Therefore, it is possible to improve the homogeneity of the optical member and, consequently, the optical performance. Further, according to the second aspect of the present invention, since the optical member has an extreme value in a region outside the effective use range, the distribution of the birefringence amount along the diameter direction of the optical member has an extreme value. Even if it does, the effect can be suppressed. (3) According to the third aspect of the present invention, with respect to the distribution of the birefringence along the radial direction of the optical member, the rate of change of the birefringence along with the change in the diameter of the optical member is 0.2 nm / cm / cm.
When the diameter is not more than cm, the rate of change of the distribution of birefringence along the diameter direction of the optical member can be reduced, and the optical performance can be improved by increasing the homogeneity of the optical member. (4) According to the invention described in claim 4, the pattern that can be projected on the substrate can be made finer.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of a synthesis furnace to which a method for manufacturing an optical member according to an embodiment of the present invention is applied.

FIG. 2 is a diagram schematically showing a configuration of a projection exposure apparatus according to an embodiment of the present invention.

FIG. 3 is a diagram showing a configuration of a projection optical system incorporated in the projection exposure apparatus according to the embodiment of the present invention.

FIG. 4 is a diagram illustrating a distribution of a birefringence amount along a radial direction of the optical member.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Burner (multi-tube burner) 2 ... Rotary target 3 ... Refractory 4 ... Furnace frame 6 ... Exhaust system 7 ... Ingot 8 ... Support shaft 10 ... Driving device 20 ... Beam shaping optical system 21 ... 1st
Relay optical systems 30, 50 Optical integrator 40 Second relay optical system 60 Condenser optical system M Mask MS Mask stage PL Projection optical system W Photosensitive substrate WS Substrate stage

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) H01L 21/027 H01L 21/30 515D F-term (Reference) 2H097 CA12 CA13 GB00 4G014 AH15 4G062 AA04 BB02 CC07 MM02 MM28 NN03 NN16 5F046 BA03 CA07 CB01 DA12

Claims (4)

[Claims] 1. A quartz glass which is spouted from said multi-tube burner while hydrolyzing a rotary target and a multi-tube burner to hydrolyze in an oxyhydrogen flame and to be vitrified on said rotary target. In the method for manufacturing an optical member, D is a value when the relative distance between the center of the multi-tube burner and the center axis of the rotary target is greatest with the swing, and Radius R
Wherein the rocking is performed so as to satisfy the following expression. R / 5 ≦ D ≦ R 2. An optical member used in a specific wavelength band of 400 nm or less, wherein a shape of a distribution curve showing a distribution of a birefringence amount along a diameter direction of the optical member is near a center of the optical member, An optical member having an extreme value in a region outside the effective use range of the optical member. 3. An optical member used in a specific wavelength band of 400 nm or less, wherein a distribution of a birefringence amount along a radial direction of the optical member is a variation of a birefringence amount with a diameter change of the optical member. The ratio is 0.
An optical member having a thickness of 2 nm / cm / cm or less. 4. A projection exposure apparatus for projecting a pattern formed on a mask onto a substrate by a projection optical system, wherein the optical member according to claim 2 is used. Projection exposure equipment.