JP2002257523A - Ultraprecise shape measuring method and its device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an utraprecise shape measuring method and its device capable of measuring highly precisely the measurement surface shape having an optional surface shape by utilizing stability of an optical path without using a reference plane used in, for example, an optical interference method, and miniaturizing the device. SOLUTION: After setting respective standard positions of a measuring system 4 and a measuring object 8, each point on a measurement surface is irradiated with convergent light emitted from a light source of the measuring system. Both positions and angles of the measuring system and the measuring object are finely adjusted so that optical axes of incident light and reflected light on the point are overlapped each other. The distance from the light source to the measurement surface is measured, and normal vectors at each point on the measurement surface are measured from deviations of the position and the angle from the standard positions of the measuring system and the measuring object. Inclination on each point on the surface are calculated from the normal vectors, and the inclination at an optional point is interpolated, and the inclination is integrated over a measurement region, to thereby calculate the surface shape.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an ultra-precision shape measuring method and apparatus, and more particularly, to a surface processed with high precision including an optical element such as an aspherical high-precision mirror for synchrotron radiation or X-rays. The present invention relates to an ultra-precision shape measuring method and apparatus used for measuring and evaluating a shape or for measuring a deviation from an ideal shape when performing correction processing.

[0002]

2. Description of the Related Art Ultra-precise measurement of a surface profile machined with high precision has become increasingly important for the development of a super-precision machining method and the improvement of precision of a super-precision device. In particular, in the case of an optical element such as an aspheric high-precision mirror for synchrotron radiation or X-rays,
The deviation of the reflecting surface shape from the ideal shape greatly affects the resolution, resolution, and accuracy of the apparatus system.

Hitherto, as a method for measuring the shape of a processed surface with high precision, an optical interference method is generally used as the method with the highest accuracy. However, this optical interference method requires a reference surface to be compared, and it is impossible to measure the shape of a highly accurate surface having a shape accuracy higher than the reference surface. In a three-dimensional coordinate measuring instrument, the linear motion accuracy of the stage determines the measurement accuracy. However, since the accuracy of a linear guide is generally limited to a micron order or a submicron order, it cannot be used for shape measurement of a nanometer order. .

Recently, large accelerators for generating synchrotron radiation have been operated at various locations. Synchrotron radiation is excellent in its polarization and parallelism and has an unprecedented high intensity, especially in a very wide wavelength range, and is used for cutting-edge microfabrication and analysis in the engineering field, and is also used in the medical field. It has come to be used for. In most cases, the emitted light is not used as it is, but is controlled and used by an optical system. Therefore, in order to effectively use the radiated light, a highly accurate collecting mirror is required. For example, an ultraviolet focusing mirror is 100 mm
A concave mirror made of fused quartz having a corner and a radius of curvature of 5600 mm is required to have a surface roughness of 1 nm and a shape accuracy of 10 nm / 100 mm. However, even if a mirror of the highest standard at the present time is used, the focal position is widened to a range of 40 mm due to its shape accuracy, and the emitted light cannot be used sufficiently effectively.
For this reason, it is necessary to process the mirror with higher precision, but as a prerequisite, it is necessary to measure the shape with very high precision.

[0005] The present inventors have proposed an ultra-high precision machining method.
We are developing numerically controlled EEM (Elastic Emission Machining), and provide a method that enables processing with an accuracy of 10 nm or less over the entire processing area. In addition, a technique for measuring the shape of the ultraviolet focusing mirror with high accuracy has already been proposed. In this shape measurement method, a measurement system including a laser light source, a photodetector, and a distance measurement device is arranged at an ideal curvature radius position of a fixedly held mirror on a drive system including an XYZ stage and a two-axis goniometer, The laser light emitted from the light source is reflected by the surface of the mirror, and the reflected light is received by the photodetector, and this is repeated over each point of the mirror by changing the angle of the measurement system. At this time, the position and angle of the light source are adjusted so that the light incident on a point on the measured surface of the mirror and the reflected light at that point overlap, and the distance from the light source to the measurement point of the mirror is accurately determined. When they are matched, the normal vector at that point becomes equal to the vector of the light beam. Therefore, by accurately deriving the light vector from the adjustment amount of the drive system, the normal vector at each point of the mirror can be determined.
The slope at each point on the surface to be measured is obtained from the normal vector, the slope between them is interpolated, and then the slope is integrated to obtain the surface shape of the surface to be measured.

[0006]

However, in the shape measurement method using the normal vector described above, a high-precision spherical mirror is fixed, a measurement system is arranged near a focal point, and only the measurement system is displaced. This measures the normal vector of each point on the surface to be measured, and therefore, for the object to be measured other than the spherical mirror, the linear displacement distance of the measurement system becomes large and the position accuracy cannot be obtained, so that the accuracy is substantially high. Only the spherical mirror could be measured. Also, the radius of curvature of the spherical mirror is 56
Since the distance is 00 mm, the mirror and the measurement system must be accurately set at a position apart from each other, so that the measurement device becomes large and expansion and contraction due to a temperature change of the base table must be considered. In addition, measurement of a mirror having a radius of curvature of several km was practically impossible and was not versatile.

In view of the above situation, the present invention is intended to solve a problem in that a surface to be measured having an arbitrary surface shape is stabilized without using a reference surface such as an optical interference method. Another object of the present invention is to provide an ultra-precision shape measurement method and a device capable of performing measurement with high accuracy by utilizing the property and miniaturizing the device.

[0008]

According to the present invention, in order to solve the above-mentioned problems, after setting respective reference positions of a measuring system and an object to be measured, a convergent light emitted from a light source of the measuring system is measured. It irradiates each point on the surface, finely adjusts the position and angle of both the measurement system and the DUT so that the incident light and the optical axis of the reflected light at that point overlap, Is measured, and a normal vector at each point on the surface of the measured object is measured from the deviation of the position and angle of the measuring system and the measured object from the reference position, and at each point on the surface from the normal vector. An ultra-precision shape measurement method has been established in which the inclination is calculated, the inclination at an arbitrary point is interpolated, and the surface shape is calculated by integrating the inclination over the measurement region.

Here, in the fine adjustment of the measurement system and the object to be measured, the parallel movement drive is minimized, and the fine adjustment is performed mainly by the rotation drive so that the incident light and the optical axis of the reflected light at that point overlap. The ultra-precision shape measuring method according to claim 1, comprising:

Further, the present invention provides a measuring drive system comprising a three-axis encoder moving table and a two-axis encoder goniometer on a common base, and an object to be measured comprising at least a two-axis encoder goniometer. A drive system is disposed at an interval, and a measurement system including a light source, an optical axis position detector, and a length measuring device for measuring a distance from the light source to the surface to be measured is held in the measurement-side drive system, Further, the object to be measured is held in the object side drive system via a sample holder, and convergent light emitted from the light source is irradiated to each point on the surface to be measured, and the incident light and the reflection at that point are irradiated. The position and angle of both the measurement system and the device under test are finely adjusted so that the optical axis of the light overlaps. And the light on the surface to be measured Calculate the incident position, find the normal vector at each point on the surface to be measured, calculate the slope at each point on the surface from the normal vector, interpolate the slope at any point, and measure the slope An ultra-precision shape measuring device was constructed by calculating the surface shape by integrating over the region.

Here, it is preferable that a minute displacement of the optical axis is detected by using a four-division photodiode as the optical axis position detector.

The length measuring device may include a half mirror for changing a direction by dividing reflected light, a cylindrical lens,
The distance from the light source to the surface to be measured is determined by using the characteristic that the spot shape of the surface orthogonal to the optical axis of the light passing through the cylindrical lens changes along the optical axis. It is also preferable that each of the driving systems is adjusted so as to be constant.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, embodiments of the present invention will be described in more detail with reference to the accompanying drawings. The present invention utilizes the optical path stability of light passing through a medium having a uniform refractive index, and adjusts a normal vector of each point on the surface to be measured such that incident light and reflected light at each point overlap. This is a method of measuring the surface shape with high accuracy by calculating the inclination of the surface at each point, interpolating and integrating the inclination.

A shape measuring apparatus according to the present invention will be described below with reference to FIGS. This shape measuring apparatus is provided with a 5-axis measuring-side driving system 2 and at least a 2-axis device-to-be-measured object driving system 3 on a common base 1. A measurement system 4 having a position detector 7 and a length measuring device 9 for measuring a distance from the light source 6 to the object 8 is held. The object side drive system 3 is connected to the object side drive system 3 via a sample holder 5. The measurement object 8 is held. Measurement side drive system 2
Is a three-axis moving table with encoder 10 (x, y,
z) and a goniometer 11 with a two-axis encoder (θ,
φ). Further, the DUT side drive system 3
Similarly, the goniometer 12 with a two-axis encoder (α,
β). In the present embodiment, the goniometer 11 is provided on a three-axis encoder-equipped moving table similar to the above in order to facilitate the initial setting of the DUT 8.

He-Ne laser 13 of transverse single mode oscillation
After passing through the optical isolator, the laser light emitted from the optical fiber 14 was guided by the optical fiber 14 and condensed by the collimator lens 15 (opening angle 23 °). Here, the coordinate system of the normal vector measurement has a detector 7 as an origin and φ, Z around the X axis.
The circumference of the axis was defined as θ. The pinhole 16, the collimator lens 15, and the optical fiber 14 are X and Y by a piezo actuator and a micrometer head via a three-point support adjusting mechanism to make the Z axis direction coincide with the optical axis direction.
It is fixed on a two-axis table capable of fine adjustment of 0.1 μm or less on a plane (not shown). Further, a guide rail can be provided so that the entire X and Y stages can be moved up and down for adjusting the pinhole position in the Z direction, and can be manually fixed at an arbitrary position (not shown). The beam splitter 17 and the ４ wavelength plate 18 were integrated, and a double V groove was used as a guide so that the beam splitter 17 could be moved accurately in the Y direction. Also, 1 /
The four-wavelength plate 18 fixes the deflection surface at a position rotated by 45 °, and converts linearly polarized light into circularly polarized light.

The light emitted from the pinhole 16 is bent by 90 ° by the beam splitter 17, passes through the 波長 wavelength plate 18, and then passes through the condenser lens 19. The condensing lens 19 is provided with a guide and a scale so that it can be adjusted in the Y direction, and can be easily positioned. Thereafter, the light is reflected by the surface to be measured 20, passes through the condenser lens 19, travels straight through the beam splitter 17, passes through the half mirror 21, and is received by the detector 7, whereby normal measurement is performed.
On the other hand, the light divided into 21 by the half mirror is received by the detector 23 through the cylindrical lens 22, and the change in the distance from the pinhole 16 to the measured surface 20 is measured. The pinhole 16 and the detector 7 detect the surface 2 to be measured.
The optical distances to 0 are exactly matched. here,
The optical axis position detector 7 and the detector 23 use a quadrant photodiode (QPD).

A detector 23 for measuring a change in the distance from the detector 7 to the surface 20 to be measured is attached to the measuring system 4 with the optical axis oriented in the Z-axis direction. Beam splitter 17
Is split 1: 1 by the half mirror 21 to the detector 7 and to the detector 23 by 1: 1. The beam split to the detector 23 side is the cylindrical lens 2
2 and is imaged on the detector 23. The detector 23 is
X, Y, having the same mechanism as the adjustment of the pinhole 16
It is mounted on a holder integrated with a mechanism that can be adjusted in the Z direction. Adjustment of the detector 23 in the Z direction can be performed by a linear guide and a micrometer head. That is, the length measuring device 9 includes the detector 23 and a mechanism for finely adjusting the detector 23.

The shape measuring apparatus according to the present invention utilizes the stability of an optical path in a uniform medium having a uniform refractive index.
Except for the laser controller 24, the whole is placed in a constant temperature chamber 25, the internal temperature is kept at ± 0.5 ° C. or less, and the temperature stability of 0.02 ° C. or less is secured over the optical path. Furthermore,
In the present embodiment, a copper pipe 26 having high thermal conductivity is arranged along the optical path.

Next, the principle of shape measurement according to the present invention will be described.
First, the light beam emitted from the light source is reflected at a certain point on the surface 20 to be measured, and the position of the light source 6 is detected using a QPD (detector 7) so that the optical axis of the incident light and the reflected light at that point overlap. When (x, y, z) and the angle (θ, φ) are adjusted, the normal vector of the measured surface 20 at that point coincides with the vector of the incident light beam. The detector 7 at this time is set as the origin, and the point on the surface 20 to be measured where the light beam enters is set as the reference position. Next, the angle of the measurement system 4 is slightly changed to slightly shift the position of the point on the measured surface 20 where the light beam enters. in this case,
The reflected light reflected by the measured surface 20 does not overlap with the incident light,
Since there is a slight shift, adjustment is made again so that the optical axes of the incident light and the reflected light overlap, and a normal vector at a new point is obtained from the adjusted (x, y, z) and (θ, φ) values. . This operation is repeatedly performed over the measurement area, and the normal vector at each point on the measurement target surface 20 is measured. If the distance between the light source 6 and the surface to be measured 20 has been measured, the inclination at each point on the surface to be measured 20 is obtained from the normal vector, and after interpolating with an appropriate function, integration over the measurement region is performed. Can be used to measure the surface shape.

More specifically, as shown in FIG.
(0, 0, 0) is the first measurement point, and the Y axis of the position adjusting coordinate system of the light source 6 (actually, the detector 7) when the normal vector and the light vector match at that time are matched, The value is O (x ₀ , y ₀ , z ₀ ), and the angles about the X axis and the Z axis are (0, 0). O ′ (x ₁ , y ₁ , z ₁ ) and (θ ₁ , φ ₁ ) in the measurement at the point P ′ (x, y, z) represent the light source position and angle when the respective vectors are matched. ing. In this way, the surface inclination and the light source position at each point on the measured surface 20 are obtained. However, since the slope of this surface is measured only at a specific point, we can use it as an appropriate function,
For example, by interpolating with a spline function, the inclination of the surface at an arbitrary point can be calculated, and by integrating the inclination, the surface shape of the measured surface 20 is calculated and obtained by an arithmetic device such as a personal computer. . Here, if the distance from the light source 6 (detector 7) to the measured surface 20 has been measured, or the distance from the light source 6 (detector 7) to the measured surface 20 when the reference coordinates are determined, By keeping the measurement constant, the surface 20 to be measured irradiated with the light beam
The coordinates of the upper measurement point can be specified. In the present invention, the requirement for the measurement accuracy of the position of the measurement point on the surface to be measured 20 may be relatively low, and it can be implemented without any problem if 1 μm or less can be secured. The accuracy of each axis of the goniometer is 1.8 × 1
It has been confirmed that it has an absolute accuracy of 0 ^-7 rad or less.

FIG. 3 shows an optical system for measuring a normal vector of the surface 20 to be measured. The coordinates in the measurement system are such that the beam traveling direction is the Y axis, and the horizontal and vertical directions are the X and Z axes. The linearly polarized parallel light passes through the pinhole 16 to cause Fraunhofer diffraction to spread the light, which is then bent by 90 ° by the beam splitter 17 and converted to a circularly polarized light by a quarter-wave plate. Through 18, the light reaches a condenser lens 19. The distance from the pinhole 16 to the condenser lens 19 is defined as a '. The light exiting the condenser lens 19 is arranged so that an image is formed at the optical path length b + 2c, and the measured surface 20 is b +
Place at the location of c. The light reflected by the surface to be measured 20 forms an image once at the point c, and is again converted into linearly polarized light by the same condensing lens 19 through the quarter-wave plate 18, and the wavefront is rotated by 90 ° to form a beam. The splitter 17 goes straight and forms an image on the optical position detector 7. As shown in FIG. 4, the detector 7 is constituted by four-division photodiodes, and divided cells 7A to 7D are arranged in each quadrant of the X and Z coordinates. When the normal vector of the surface to be measured 20 changes, the position of the pinhole image on the detector 7 is displaced by the principle of optical leverage. Each of the divided cells 7A to 7A according to the displacement amount of the pinhole image in the X and Z directions.
Output varies from _{D V A, V B, V} C, appear as V _D, V by a respective subtraction horizontal, as the position change amount in the vertical direction
_{1 = (V A + V B} ) - (V C + V D), V 2 = (V A + V D) -
(V _B + V _C ) is obtained. The relationship between the amount of change in the normal vector and the amount of positional displacement on the detector 7 is determined by the values of a ', a, b, c. The measurement resolution of the normal vector is determined by the position displacement detection sensitivity of the detector 7. Detector 7
When measuring a surface having undulation or curvature of the measured surface 20 using only a measuring system, the distance from the pinhole 16 to the measured surface 20 changes and the lever arm changes and the image is formed. The position may change, and it may not be possible to measure an accurate normal vector. Therefore, a length measuring device 9 (such as a detector 23) for accurately measuring the distance from the pinhole 16 to the surface 20 to be measured is used.

In order to accurately measure the distance from the pinhole 16 to the surface 20 to be measured, it is sufficient to use an optical system which changes the state of some light for measuring the imaging position.
FIG. 5 shows an optical system for measuring the imaging position.
As shown in FIG. 3, the light is split by placing a half mirror 21 in front of the detector 7, and an image is formed using a cylindrical lens 22. As shown in FIG. 5A, on the detector 23 composed of a four-division photodiode, if the distance to the surface to be measured 20 is short, the light cross-sectional shape becomes an elliptical shape that is long in the vertical direction, and the distance is long. If this is the case, it becomes an elliptical shape that is long in the horizontal direction (see FIG. 5B), and the change in the cross-sectional shape of light, that is, the change in the imaging position is measured as the change in V ₃ and V _{4 in} the same manner as the detector 7. be able to. In the case of this detector 23, the arrangement of the four-division photodiodes must be inclined by 45 ° with respect to the X and Z axes.

The resolution of the imaging position is determined by the relationship shown in FIG.
The focal length of the condensing lens 19 is f ₁, Cylindrical
Focal length by lens 22 and f₁Distance from focus by
F_Two, H, to the cylindrical lens 22
The focal distance e through the cylindrical lens 22
F from where the light becomes a perfect circle₁Distance to focal position
Let g be the following equation (1).

[0024]

(Equation 1)

Δe when b changes by Δb,
Δh is expressed by the following equation (2).

[0026]

(Equation 2)

When b is displaced by Δb, the detector 23
Above, the image is not a perfect circle but an ellipse. a direction that is bent by f _2, the optical length ratio of the shape of the direction that is not applied to f ₂ if K, is expressed by the following equation (3).

[0028]

(Equation 3)

The change in shape on the detector 23, ΔK / Δh, represents the sensitivity of the imaging position measurement, and the sensitivity can be increased by reducing f ₂ .

According to the present invention, the shape of the surface is obtained by measuring the direction of the normal line of an arbitrary point on the surface to be measured with high precision, calculating the slope, interpolating, and then integrating. Therefore, the shape of the surface to be measured is an arbitrary shape, for example, a concave surface, a convex surface,
It can be measured not only in a shape represented by a mathematical expression such as an aspherical surface, a spheroid, or a toroidal shape, but also in any shape and regardless of the size of the surface to be measured. In addition, this method does not use any artificially created reference surface or the like. Since the measurement accuracy in the direction of the normal line is dominant in this measurement accuracy, by improving the measurement accuracy in the direction of the normal line, a diameter of 10
A free-form surface of 0 mm can be measured with an accuracy of 1 nm or less.

In order to make this measurement method a practical method, the stability of the optical path of light in a space where the refractive index does not fluctuate is used to convert a light beam emitted from an arbitrary coordinate point to an arbitrary surface of the object to be measured. By adjusting the two-axis goniometer and the three-axis moving table so that the reflected light is located at the same position as the emitted coordinate point.
The normal at that point can be measured. If the direction of the normal line is almost parallel to a certain direction, for example, when measuring a flat surface in the shape of the object to be measured, any one of the three axes of the moving table is set to the length of the object to be measured. Must be moved. In this case, the accuracy of the moving table deteriorates the measurement accuracy. Geometrically movable tables cannot be manufactured.

Therefore, in order to minimize the amount of movement of the moving table when the incident and reflected optical paths are made the same, if the object to be measured is biaxially rotated on a plane perpendicular to the optical path, high accuracy can be achieved. In addition, by making the radius of curvature of the rotation axis and the object to be measured close to each other according to the curvature of the surface of the object, it is possible to achieve the same incident and reflected light paths only by rotating the goniometer of a total of four axes of the measurement system and the sample system. It is possible to do. In other words, rotary drive can be controlled with much higher accuracy than linear drive,
It is the basic principle of the present invention to utilize this to increase the measurement accuracy.

According to the present invention, the normal of an arbitrary point on the surface of an object to be measured is measured very precisely by utilizing the stability of light and an optical path without using an artificial reference plane. This is a method for measuring any shape ultra-precisely by calculating, interpolating and integrating the slope of. In the ultra-precise measurement of the normal, it is also a feature of the present invention that the normal is measured by the same optical path of incident and reflected light. This measurement method can be achieved by using a 4-axis goniometer and a 3-axis moving table.

Next, the procedure for measuring various shapes will be briefly described.

(Measurement of Shape of Spherical Surface and Concave Surface) First, a sample is set on a holder, and initial coordinates, that is, an origin is determined. Then, the light source is arranged at the center of curvature of the spherical surface which is the surface to be measured. The light beam emitted from the light source is reflected at the origin of the surface to be measured, and the two-axis goniometer (θ, φ) and the three-axis moving table (x, y, z).

Next, a light beam from the light source is emitted to a certain measuring point, and the two-axis goniometer (θ, φ) and the three-axis movement are adjusted so that the normal vector of the surface to be measured at that time coincides with the light vector. Adjust table (x, y, z). At this time,
The detector 23 is arranged so that the distance between the surface to be measured and the light source does not change.
The detector 7 is used to adjust the deviation between the normal vector and the light beam vector (identify the incident and reflected light). The position (x, y, z) and angle (θ,
From the value of φ), a normal vector is obtained.

The above operation is repeated to measure the normal vector at an arbitrary point on the measured surface, calculate the inclination of the measured surface at that point, and interpolate and integrate the inclination to obtain the surface shape. Ask for it.

FIG. 7 shows a case where a spherical mirror is selected as the object 8 to be measured. FIG. 7 (a) shows an example in which initial coordinates are set and a reference between the light source 6 (detector 7) of the measuring system and the surface 20 to be measured. The position has been determined. Then, as shown in FIG.
When the angle of the incident light emitted from the light source 6 is changed by Δθ, the spot image of the reflected light is displaced on the detector 23. While displacing the light source 6 by Δx in the X-axis direction,
The θ axis is finely adjusted so that an image is formed at the center of the detector 23. In this case, since the measurement target surface 20 is a spherical surface, if the detector 23 is arranged at the center of the curvature, the detector 23
Is very small, the error due to the linear guide is small, and the normal vector can be measured with high accuracy.

(Measurement of Shape of Plane and Aspherical Surface) As in the case of the above-mentioned spherical surface, a sample is set on a holder and initial coordinates, that is, an origin is determined.

Next, the normal vector at an arbitrary point on the surface to be measured can be measured in the same manner as a spherical surface, but the moving amount of the moving table (x, y, z) becomes large, and the parallel moving table becomes large. The motion accuracy of the image makes the measurement accuracy of the normal vector worse. Therefore, in order to minimize the amount of parallel movement, if a two-axis goniometer (α, β) perpendicular to the optical axis is arranged in the device to be measured and the sample is set in a holder provided thereon, Accurate measurement of the normal vector can be realized. This is because rotation is easier to achieve accuracy than translation.

In the case of an object to be measured having an arbitrary curvature, the distance between the rotation axis and the surface to be measured is made closer to the curvature of the surface to be measured according to the curvature of the surface to be measured, so that the translation table can be moved. (Θ, φ), (α, β) 4 with almost no movement
The rotation of the axial goniometer alone makes the incident and reflected optical paths coincide and the normal vector can be measured with high accuracy. Others are the same as above.

FIG. 8 shows a case where a plane mirror is selected as the object 8 to be measured, and FIG. 8A shows a state in which initial coordinates are determined. FIG. 8B shows a case where, when irradiating different points on the surface to be measured 20, only the measurement system is changed in the X-axis direction while the object to be measured 8 is fixed.
In this case, if an error occurs in the X-axis linear guide,
The incident light emitted from the light source 6 (detector 7) is
And the reflected light is directed in a direction different from that of the incident light, so that it is accurately formed at the center of the detector 7 by θ.
It becomes to adjusting the axial only theta _err, or the theta _err that such by deviation of the measurement surface 20 can not determine whether something's by the linear guide of the error in the X-axis. Therefore, in the present invention, first, the measurement system adjusts the θ-axis of the measurement-side drive system 2 and adjusts the DUT 8 to adjust the α-axis of the DUT-side drive system 3. , The reflected light from the surface to be measured 20 can be substantially accurately detected.
And the slight deviation from the center is adjusted along the XYZ-axis linear guide. In this case, since the amount of parallel movement in the adjustment of the measurement system is extremely small, the normal vector can be measured with high accuracy.
Similarly, the present invention can measure not only a plane but also an aspherical surface, an elliptical surface, and the shape of an object to be measured having an arbitrary surface with high accuracy.

FIG. 9 shows the result of actually measuring the shape of a spherical mirror processed with high precision. The results in FIG. 9 are obtained by repeating the same measurement four times for the same object, and have a shape error of 10 nm and a slope error of 2 × 10 ⁻⁷ ra.
This indicates that reproducibility of d or less was achieved. Also,
FIG. 10 shows the results of measuring the AA line and the BB line on the sample surface with the shape measuring apparatus of the present invention to obtain a shape error from an ideal curved surface (0 level) (before processing), and processing the surface based on the error. The figure also shows the results (after EEM processing) of measuring the shape error from the ideal curved surface after processing by numerical control EEM in accordance with the obtained amount. According to the present invention, not only can the surface shape be measured with high accuracy, but also data on the shape error from the ideal curved surface at each measurement point can be obtained, and this data can be used for ultra-precision machining. Further, since the present invention directly measures a normal vector required for ray tracing on a mirror, it can be used as data for calculating the convergence of emitted light by a converging mirror, as well as shape accuracy. It is a big feature.

Lastly, in this embodiment, a sample holder 5 capable of holding an object to be measured without changing its original shape will be described with reference to FIGS. When holding the device under test, if local stress is applied to the device under test,
The distortion due to the stress appears as a change in surface shape. Since the present invention has such an accuracy as to measure even such a slight change in shape, it is held so as not to cause a change in shape on the surface of the object to be measured. It must be held so that it does not move. FIG.
FIG. 12A shows a state in which the sample holder 5 is attached to the device-to-be-measured drive system 3, and FIG. 12A is a plan view of the sample holder 5.
(B) shows a sectional view thereof. The sample holder 5 is air-tightly joined to an aluminum holding frame 27 via an O-ring 30 with a 10 mm thick aluminum surface plate 28 having a recess 29 formed in a stepped shape while leaving the periphery on the back surface. Has a hole 31 of 4 mm pitch and a diameter of 1 mm in 100 × 300
mm, and a space formed between the holding frame 27 and the surface plate 28 by a hose 32 having a diameter of 6 mm can be evacuated by a diaphragm type vacuum pump.
Has a structure in which a 500 μm-thick Teflon (registered trademark) sheet 33 having micro continuous holes is laminated on the surface thereof.
Here, the Teflon sheet 32 used was a trade name “Gore-Tex” (trademark of WL Gore & Associates, USA). This Gore-Tex is a breathable material, but has fine pores that do not allow water to pass through, and is also used as an elastic gasket.

Then, with the back surface of the object to be measured joined to the Teflon sheet 33, the inside is evacuated and the front surface of the back surface of the object to be measured is suctioned and held with a uniform suction force through the Teflon sheet 33. is there. When the object to be measured is allowed to stand on a flat surface and when the object is held vertically by the above-described sample holder 5,
As a result of observing the surface state with an interferometer (Zygo), neither interference fringe could be distinguished at all. This means that the object to be measured can be held by the sample holder 5 without affecting the surface shape at all. Also, an aluminum plate (50 × 50
mm), a copper mirror (80 × 250 mm, thickness 25 mm) with a chuck surface serving as a rough surface, and a glass (BK7) flat mirror (50 × 200 mm, thickness 30 mm) with a sand surface on the chuck surface. Was adsorbed,
It could be held without any problems. Incidentally, the above-mentioned copper mirror and plane mirror cannot be fixed at all by a normal vacuum chuck.

[0046]

According to the ultra-precision shape measuring method and apparatus of the present invention as described above, not only the spherical concave shape, but also
A surface to be measured having an arbitrary surface shape as well as a shape represented by a mathematical expression such as a convex surface, an aspheric surface, a spheroid, and a toroidal shape, which could not be optically measured with an accuracy of 1 nm or less. The shape can be measured with high accuracy using the stability of the optical path without using a reference surface such as optical interferometry, and the surface shape can be reduced in a short time regardless of the size of the surface to be measured. It has an excellent effect that measurement can be performed and the size of the measurement device can be further reduced.

Further, according to the present invention, not only can the surface shape be measured with high accuracy, but also data on the shape error from the ideal curved surface at each measurement point can be obtained, and this data can be used for ultra-precision machining. Further, since the present invention directly measures a normal vector required for ray tracing in a mirror, the present invention can be used as data for calculating the convergence of emitted light by a converging mirror, as well as in shape accuracy.

[Brief description of the drawings]

FIG. 1 is an external view of an ultra-precision shape measuring apparatus according to the present invention.

FIG. 2 is an explanatory diagram of a measurement principle of the present invention.

FIG. 3 is a simplified layout diagram of an optical system.

FIG. 4 is an explanatory diagram showing a state in which reflected light is received by an optical axis position detector using a four-division photodiode.

FIG. 5 shows the principle of distance measurement by a length measuring instrument, wherein
FIG. 3B is an explanatory view showing a state in which reflected light is received by a detector using a divided photodiode, and FIG. 4B shows a state in which the spot shape of a surface orthogonal to the optical axis of light passing through a cylindrical lens changes along the optical axis. FIG.

FIG. 6 is an arrangement diagram of an optical system showing a principle of distance measurement by the length measuring device.

FIG. 7 shows a procedure for measuring the surface shape of a spherical mirror,
(A) is an explanatory diagram of a state in which an initial position is set, and (b) is an explanatory diagram showing a state in which reflected light shifts due to a deviation from an ideal curved surface when the direction of a light source is changed.

FIG. 8 shows a procedure for measuring a planar surface shape, and FIG.
Is an explanatory view of a state in which an initial position is set, and (b) shows a relationship between a state of incident light and reflected light and a translation amount of the measurement system when the object to be measured is fixed and the direction of the light source is changed. FIG. 3C is an explanatory diagram showing the relationship between the state of incident light and reflected light and the amount of parallel movement of the measurement system when the object to be measured is rotated and the direction of the light source is changed.

FIG. 9 is a graph showing a result of measuring a spherical shape.

FIG. 10 shows a sample surface measured by the present shape measuring apparatus to determine a shape error from an ideal curved surface.
9 is a graph showing a result of measuring a shape error from an ideal curved surface as a result of processing.

FIG. 11 is a side view showing details of a device-side drive system and a sample holder.

FIG. 12 shows a sample holder, (a) is a front view,
(B) shows a sectional view.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Base stand 2 Measurement-side drive system 3 Object-side drive system 4 Measurement system 5 Sample holder 6 Light source 7 Optical axis position detector 8 Object to be measured 9 Length measuring device 10 Moving table with 3-axis encoder 11 2-axis encoder Goniometer with goniometer 12 Goniometer with 2-axis encoder 13 He-Ne laser 14 Optical fiber 15 Collimator lens 16 Pinhole 17 Beam splitter 18 Quarter-wave plate 19 Condenser lens 20 Measurement surface 21 Half mirror 22 Cylindrical lens 23 Detector 24 Controller 25 Constant temperature chamber 26 Copper pipe 27 Holding frame 28 Surface plate 29 Depression 30 O-ring 31 Hole 32 Hose 33 Teflon sheet

Claims (5)

[The claims] After setting respective reference positions of a measurement system and an object to be measured, convergent light emitted from a light source of the measurement system is irradiated to each point on a surface to be measured, and the incident light and the point Fine-tune the position and angle of both the measurement system and the DUT so that the optical axis of the reflected light overlaps, measure the distance from the light source to the DUT, and determine the reference position of the measurement system and the DUT. Measure the normal vector at each point on the surface of the measured object from the deviation of the position and angle from, and calculate the slope at each point on the surface from the normal vector, and interpolate the slope at any point, An ultra-precision shape measurement method characterized by calculating a surface shape by integrating the inclination over a measurement area. 2. In the fine adjustment of the measurement system and the object to be measured, parallel movement driving is minimized, and fine adjustment is performed mainly by rotational driving so that the incident light and the optical axis of the reflected light at that point overlap. The method for measuring ultra-precision shape according to claim 1. 3. A measurement-side drive system including a three-axis encoder goniometer and a two-axis encoder goniometer on a common base table, and a DUT-side drive system including at least a two-axis encoder goniometer. The measurement side drive system is provided with an interval, and a measurement system including a light source, an optical axis position detector, and a length measuring device for measuring a distance from the light source to a surface to be measured is held, and The object to be measured is held in the object side drive system via the sample holder, and the convergent light emitted from the light source is irradiated to each point on the surface to be measured, and the optical axis of the incident light and the reflected light at that point The position and angle of both the measurement system and the device under test are finely adjusted so as to overlap with each other, and the arithmetic unit having the control function of each drive system obtains the deviation of each axis from the reference position from each drive system. And calculate the incident position of light on the surface to be measured Then, a normal vector at each point on the surface to be measured is obtained, a tilt at each point on the surface is calculated from the normal vector, a tilt at an arbitrary point is interpolated, and the tilt is integrated over the measurement area. An ultra-precision shape measuring device characterized in that the surface shape is calculated by calculation. 4. The ultra-precision shape measuring apparatus according to claim 3, wherein a minute displacement of an optical axis is detected by using a four-division photodiode as the optical axis position detector. 5. The length measuring device is composed of a half mirror that changes the direction by dividing reflected light, a cylindrical lens, and a detector including a four-division photodiode, and is arranged on the optical axis of the light passing through the cylindrical lens. 5. The drive system according to claim 3, wherein the drive systems are adjusted so that the distance from the light source to the surface to be measured is constant using a characteristic in which the spot shape of the orthogonal surface changes along the optical axis. Ultra-precision shape measuring device.