JP2022162306A - Surface shape measurement device and surface shape measurement method - Google Patents

Abstract

To provide a surface shape measurement device and surface shape measurement method that enable improvement in measurement accuracy of a surface shape of a measured object having a measured plane without requiring a physical reference plane, and enable a measurement of a thickness and parallelism thereof.SOLUTION: A surface shape measurement device is configured to: have a converging point PQ of spherical wave illumination light Q and a converging point PL of inline spherical wave reference light L set at a mutually mirror image arrangement with respect to a virtual plane VP, detect a position of the measured plane so that a measured plane contacts with the virtual plane VP, and make a movement adjustment of a position of a sample stand holding a measured object with a bottom reference; record each hologram of object light O being reflection light of the illumination light Q with which the measured plane is obliquely irradiated and the inline spherical wave reference light L; generate, in the virtual plane VP, a measurement-purpose reproduction object light hologram hV and a spherical wave light hologram sV representing spherical wave light to be discharged from a mirror image point of the spherical wave illumination light Q; obtain a height distribution of the measured plane from a phase difference distribution of both holograms hV and sV; and obtain a thickness of the measured plane from an amount of movement of the sample stand.SELECTED DRAWING: Figure 1

Description

The present invention relates to a surface shape measuring device and a surface shape measuring method in digital holography.

Conventionally, digital holography, which acquires the intensity and phase of light waves as digital data or generates and analyzes holograms on a computer, is a technology for analyzing light waves such as reflected light and transmitted light.

In digital holography, various techniques have been proposed to achieve high-speed acquisition and high-precision processing of hologram data. For example, digital holography is known that applies spatial frequency filtering and spatial heterodyne modulation to hologram data recorded in one shot to quickly and accurately generate a complex amplitude in-line hologram for object image reconstruction (see, for example, patent Reference 1).

In order to solve the problems of conventional optical microscopes, holography is used to accurately record a large numerical aperture object beam in one shot without using an imaging lens, and to enhance the recorded object beam by plane wave expansion. A method for accurately reproducing a resolution three-dimensional image by computer is known (see, for example, Patent Document 2). Since such a microscope does not use an imaging lens, it is possible to solve the problem of the medium and imaging lens that conventional optical microscopes have.

In addition, a high-resolution tomographic imaging method using a reflection-type lensless holographic microscope and a wavelength-swept laser beam is known to measure the internal structure of cells in a culture medium and biological tissue at high resolution (see, for example, patent documents 3).

Furthermore, the large numerical aperture object light emitted from an object irradiated with illumination light in different incident directions is recorded as hologram data for each incident direction of the illumination light, and these multiple large numerical aperture holograms are converted into one hologram. A method of synthesizing and reproducing object light under a synthetic numerical aperture exceeding 1 is known (see, for example, Patent Document 4). According to this method, a super-high-resolution three-dimensional microscope with a resolution exceeding the usual diffraction limit can be realized.

In addition, there is known a holographic ellipsometry device that uses accurate recording of light waves by one-shot digital holography and plane wave expansion of the recorded light waves (see, for example, Patent Document 5). According to this ellipsometry device, the data of the reflected light from the incident light having a large number of incident angles, including the non-parallel illumination light, can be collectively recorded in the hologram. The metric angles Ψ and Δ can be obtained, and the measurement efficiency can be improved.

In addition, an image sensor, two imaging lenses, a cube beam splitter, an optical element having a Fizeau reference plane, and an object to be measured are arranged in series, and interference between two reflected lights from the reference plane and the part to be measured is detected. A measuring device that records fringes and performs shape measurement is known (see Patent Document 6, for example).

Holography as shown in these Patent Documents 1 to 5 is applied to microscopic observation and shape measurement of a relatively small area. There is a demand to respond to

In addition, the measuring apparatus shown in the above-mentioned Patent Document 6 uses Fizeau interference, which is a general method for measuring flatness. has the following inherent problems.

The Fizeau interferometer is considered to be one of the devices that can measure flatness with the highest accuracy and speed, and is used as a flatness measurement device in standard instrument laboratories around the world. In the Fizeau interference measurement, the interference fringes formed by the light reflected by the reference plane of the transparent glass plate and the light reflected by the surface to be measured are recorded. In order to improve the measurement accuracy, the reference plane is slightly shifted in the vertical direction to shift the phase of the interference fringes, and multiple interference fringes with different phases are recorded. Analyze the planar shape of the surface. The result of such measurement is only a comparison between the reference plane and the surface to be measured, and absolute shape correction of the reference plane is necessary to measure the absolute value of the flatness. A three-sheet combination method is used for absolute shape correction.

The optical system of the Fizeau interferometer has a relatively small number of optical parts and a simple structure. A turntable or the like is required for absolute shape correction. Measurement accuracy is affected by uncertainties in reference plane shape correction, uncertainties in phase shift, and uncertainties due to environmental fluctuations. It is difficult to suppress the uncertainty of the combined measurement to 10 nm or less. Another problem is that the use of a reference plane and a collimator lens limits the diameter of the object to be measured to approximately 300 mm or less, and it is difficult to increase the diameter beyond that. In addition, there is a problem that the contrast of the interference fringes is lowered for the surface to be measured whose reflectance is greatly different from that of the reference plane made of glass, making highly accurate measurement difficult.

Therefore, in order to solve the above-mentioned demands and problems, a surface shape measuring apparatus and a surface shape that can improve measurement accuracy without using a physical reference plane as a comparison target for shape measurement and without relying on a mechanical adjustment mechanism. A measurement method has been proposed (see Patent Document 7, for example).

The apparatus and method disclosed in this patent document 7 correspond to the height from the virtual plane set at the position of the surface to be measured using a hologram that records the reflected light of the spherical wave illumination light from the surface to be measured. Then, the phase of the light wave is obtained, the phase of the spherical wave emitted from the point light source of the spherical wave illumination light calculated on the virtual plane is obtained as the phase of the reference plane, and the phase difference between the two is obtained. From the distribution of the phase difference on the virtual plane and the information on the wavelength, the height distribution on the virtual plane is obtained as the measured value of the surface shape.

Therefore, according to the apparatus and method disclosed in Patent Document 7, the phase data of the reflected light of the spherical wave illumination light on the surface to be measured is acquired, and the analytically obtained phase distribution of the spherical wave on the plane cross section and the Since shape measurement is performed by comparison, a physical reference plane such as a glass substrate is not required, and a large area can be handled, and highly accurate surface shape measurement can be realized.

WO2011/089820 WO2012/005315 WO2014/054776 WO2015/064088 WO2018/038064 U.S. Pat. No. 8,269,981 WO2020/045589

The surface profile measurement apparatus and method disclosed in the above-mentioned Patent Document 7 can measure the height of the surface to be measured that exists within a predetermined distance from a virtual plane that is virtually set in the optical system of the measurement apparatus. be. The predetermined distance is the wavelength when one single-wavelength light is used, and the combined wavelength when a plurality of different single-wavelength lights are used.

Therefore, in Patent Document 7, as a method of holding an object to be measured on a sample stage, the position of the sample stage is adjusted in advance so that the mounting surface of the sample stage coincides with a virtual plane, and the object to be measured is placed on the mounting surface. A method of holding an object to be measured by bringing the surfaces to be measured of the object into contact with each other is exemplified. The mounting surface is used as a reference surface to have physical elements on the virtual plane. This holding method differs from the method in which the bottom surface (or the rear surface, or the rear surface) of the object to be measured faces the surface to be measured, and is in contact with the mounting surface of the sample table. Since the device is placed in direct contact with the placement surface, setting for holding the object to be measured is easy.

However, depending on the shape of the object to be measured, it may not be possible to hold the surface to be measured so as to face and contact the mounting surface. In addition to measuring the surface shape with high accuracy, the parallelism of the surface to be measured with respect to the bottom surface can be measured based on the bottom surface of the object to be measured. Or it cannot respond to applications and requests such as measuring convex shapes.

The present invention is intended to solve the above problems, and can improve the measurement accuracy of the surface shape of the object to be measured without comparing it with a physical reference plane, and can measure the bottom surface of the object to be measured as a reference. To provide a surface shape measuring device and a surface shape measuring method capable of measuring the parallelism of a surface to be measured with respect to the bottom surface, and measuring the thickness, inclination angle, concave surface shape or convex surface shape of the object to be measured. intended to

In order to achieve the above object, the surface shape measuring apparatus of the present invention includes:
In a surface shape measuring device using holography,
Two lights: an object light (O) that is reflected light of the spherical wave illumination light (Q) that illuminates the surface to be measured of the object, and an inline spherical wave reference light (L) that is inline with the object light (O). as an object beam off-axis hologram (I _OR ) and a reference beam off-axis hologram (I _LR ) using the off-axis reference beam (R), respectively;
an image reproduction unit that performs calculation processing for reproducing the image of the surface to be measured using the data acquired by the data acquisition unit and acquiring surface shape data of the surface to be measured;
The data acquisition unit
an image sensor that converts light intensity into an electrical signal and outputs it as hologram data;
a sample table that holds the object with the surface to be measured facing the light receiving surface of the image sensor;
An illumination light condensing point (P _Q ), which is the condensing point of the spherical wave illumination light (Q), and a reference light condensing point (P _L ), which is the condensing point of the inline spherical wave reference light (L) , are mirror images of each other with respect to a virtually set virtual plane (VP), and the inline spherical wave reference light (L) passes through the virtual plane (VP) and is incident on the image sensor. an optical system for acquiring a hologram;
a position detection unit that detects the position of a point on the surface to be measured in an area including the front and back of the virtual plane (VP) in the optical axis direction of the image sensor and outputs the information as reflection position information;
a position adjustment device that moves the sample stage so that the detected point is included in the virtual plane (VP) based on the reflection position information;
The image reproducing unit
Using the fact that the light emitted from the reference light condensing point (P _L ) is a spherical wave, the light wave of the object light (O) is obtained from the data of the two types of off-axis holograms (I _OR , I _LR ). an object beam hologram generator for generating an object beam hologram (g) representing
a reconstructed object light hologram generation unit for optically propagating and rotationally transforming the object light hologram (g) to generate a reconstructed object light hologram (h ^V ) on the virtual plane (VP);
a reference point detection unit that performs optical propagation conversion on the object light hologram (g) to detect a position where the object light (O) is focused and sets the position as a reference point (S1) for shape measurement;
an analysis light hologram generator that analytically generates a spherical wave light hologram (s ^V ), which is a hologram in the virtual plane (VP) of spherical wave light emitted from the reference point (S1);
a shape measuring unit that obtains a height distribution of a surface to be measured of the object from a surface distribution of a phase difference between the reproduced object optical hologram (h ^V ) and the spherical wave optical hologram (s ^V ). .

In order to achieve the above object, the surface shape measuring method of the present invention comprises:
In a surface shape measurement method for measuring the shape of a surface to be measured of an object using holography,
A reference light condensing point (P _L ), which is a condensing point of the inline spherical wave reference light (L), is arranged on the optical axis of the image sensor, and the spherical wave illumination light (Q) is located at a position off the optical axis. An illumination light condensing point (P _Q ) is arranged, and a line segment connecting the reference light condensing point (P _L ) and the illumination light condensing point (P _Q ) is vertically bisected. Set a virtual plane (VP), which is the plane to
setting a reference point (P _O ) at the intersection position of the virtual plane (VP) and the optical axis;
holding a plane-parallel substrate having a reference plane on a sample stage so that the reference plane faces the light receiving surface of the image sensor;
Adjusting the sample table so that the reference plane is in contact with the virtual plane (VP);
holding the object instead of the parallel plane substrate on the sample table so that the surface to be measured faces the light receiving surface of the image sensor;
adjusting the position of the sample table so that the surface to be measured is in contact with the virtual plane (VP);
obtaining data of object light (O), which is reflected light of the spherical wave illumination light (Q) from the surface to be measured, as an object light off-axis hologram (I _OR ) using the image sensor;
Data of the in-line spherical wave reference light (L) passing through the virtual plane (VP) and incident on the image sensor in a state in which the parallel plane substrate and the object are not arranged are captured using the image sensor. Acquired as a reference beam off-axis hologram (I _LR ),
Object light representing a light wave of the object light (O) from the data of the two types of off-axis holograms (I _OR , I _LR ) by calculation processing using the inline spherical wave reference light (L) being spherical wave light generating a hologram (g);
optically propagating and rotationally transforming said object beam hologram (g) by computational processing to produce a reconstructed object beam hologram (h ^v ) in said virtual plane (VP);
Light propagation conversion is performed on the object light hologram (g) by calculation processing to detect the position where the object light (O) is condensed, and the position is converted to the illumination light condensing point with respect to the virtual plane (VP). Considered as a mirror image point of (P _Q ) and set as a reference point (S1) for shape measurement,
analytically generating a spherical wave light hologram (s ^v ) which is a hologram in said virtual plane (VP) of spherical wave light emitted from said reference point (S1);
The height distribution of the measured surface of the object is obtained from the surface distribution of the phase difference between the reproduced object optical hologram (h ^V ) and the spherical wave optical hologram (s ^V ).

According to the surface shape measuring apparatus of the present invention, since it comprises a position detecting section for obtaining information on the position of the surface to be measured of the object held on the sample stage and a position adjusting device for adjusting the position of the sample stage, According to the surface profile measuring method of the invention, the position of the object to be measured of the object held on the sample table is obtained and the position of the sample table is adjusted, so that the object to be measured can be held on the sample table with the bottom surface as the reference. can be done. Therefore, regardless of the surrounding structure of the surface to be measured of the object to be measured, the object to be measured can be held on the sample table to measure the surface profile with high precision, and to measure the thickness and parallelism.

4 is a flow chart showing a surface shape measuring method according to the first embodiment of the present invention; The conceptual diagram of the optical system for demonstrating the same measuring method. 4 is a flow chart illustrating a method for determining the position of a reference point that is a mirror image of an illumination light condensing point; (a) is a diagram for explaining the position setting of the sample stage performed using the plane-parallel substrate and autofocus technology, (b) is a diagram of the state in which the plane-parallel substrate in (a) is replaced with the object to be measured, and (c) ) is a diagram in which both surfaces of the object to be measured in (b) are not parallel and the position is adjusted by moving the sample table. The side view of the surface shape measuring device which concerns on 2nd Embodiment. The top view of the same apparatus. The side view of the surface shape measuring device which concerns on 3rd Embodiment. The side view which shows the modification of the same apparatus. The side view of the surface shape measuring device which concerns on 4th Embodiment. The top view of the same apparatus. The side view of the surface shape measuring device which concerns on 5th Embodiment. The side view of the image sensor periphery of the surface shape measuring device according to the sixth embodiment. The block block diagram of the surface shape measuring device which concerns on 7th Embodiment.

A surface profile measuring apparatus and a surface profile measuring method according to embodiments of the present invention will be described below with reference to the drawings.

(First embodiment: surface shape measuring method)
A surface shape measuring method according to the first embodiment will be described with reference to FIGS. 1 to 4. FIG. As shown in FIGS. 1 and 2, this surface shape measuring method is a method of measuring the shape of the surface to be measured of an object 4, which is the object to be measured, using holography. ) to the surface shape measurement step (#7).

In the optical system setting step (#1), the illumination light condensing point PQ, which is the condensing point of the spherical wave illumination light _Q , and the in-line spherical wave reference are arranged so as to be mirror images of each other with respect to the virtually set virtual plane VP. A reference light condensing point PL, which is the condensing point of the light _L , is set. Further, the image sensor 5 is arranged so that a straight line obliquely passing through the virtual plane _VP from the reference light condensing point PL is substantially at the center of the light receiving surface 50 of the image sensor 5 and substantially perpendicular to the light receiving surface 50, and the straight line and the virtual plane _VP . That is, the optical axis of the image sensor 5 is set as the normal to the center of the light receiving surface 50, and the reference point _PO and the reference light condensing point _PL are arranged on the optical axis. In addition, a plane-parallel substrate 40 having a reference surface 40a on the mounting surface of the sample table 7 is held in a bottom-based arrangement, and the position and angle of the sample table 7 and the optical system are adjusted so that the reference surface 40a is in contact with the virtual plane VP. is adjusted and set. When these settings are completed, the setting of the optical system including the sample stage 7 and the image sensor 5 is completed, and the reference point _PO becomes a point on the virtual plane VP and the reference plane 40a.

The bottom reference (also referred to as back reference) arrangement is such that the reference surface 40a of the parallel plane substrate 40 faces the image sensor 5, and the bottom surface of the parallel plane substrate 40 (also referred to as the opposite surface, back surface, or back surface of the reference surface 40a) faces the sample stage. In this arrangement, the plane-parallel substrate 4 is held on the sample table 7 so as to face the mounting surface of 7 . Similarly, the object 4, which is the object to be measured, is held on the sample stage 7 with the surface 4a to be measured directed toward the image sensor 5 and arranged on the basis of the bottom surface.

Under the above configuration, the optical system is set so that the image sensor 5 detects the spherical wave light L and the object light O which is the reflected light of the spherical wave light Q from the reference surface 40a of the parallel plane substrate 40. are acquired by the off-axis reference beam R. The hologram of the spherical-wave reference beam L is acquired in a state where the sample stage 7 and the parallel plane substrate 40 are not on the optical path of the spherical-wave reference beam L. FIG. By reproducing the hologram of each light source on a computer, adjustment setting is performed while confirming the success or failure of the setting. Each of the spherical wave lights Q, L and the off-axis reference light R are mutually coherent laser lights emitted from one light source.

For the plane-parallel substrate 40, for example, a JIS-standard optical parallel substrate having submicron accuracy in flatness and parallelism can be used. The positions of the condensing points P _Q and P _L , that is, the positions of the light sources of the spherical wave lights Q and L are set, for example, by the pinhole positions of the pinhole plate. In addition, the accuracy required for adjusting and setting the sample table 7 so as to place the reference plane 40a at the position of the virtual plane VP can be achieved with an accuracy that can be adjusted by mechanical operations such as screws. . This is because, in this surface shape measuring method, a process for increasing the measurement accuracy of the surface shape measurement to the nm (nanometer) order is performed as post-processing in the computer during image reproduction.

In the object setting step (#2), the plane-parallel substrate 40 is removed from the sample table 7, the surface 4a to be measured of the object 4 is viewed from the illumination light condensing point _PQ , and the object 4 is arranged on the sample table with the bottom surface as the reference. Hold at 7. The object 4 is generally of a different thickness than the plane-parallel substrate 40, such as, for example, silicon substrates in the semiconductor industry. Therefore, the position of the sample table 7 is adjusted using a position adjusting device for moving the sample table 7 so that the surface 4a to be measured of the object 4 is in contact with the virtual plane VP. The movement of the sample table 7 is movement toward or away from the virtual plane VP, and movement in the vertical direction to the virtual plane VP.

Bringing the surface 4a to be measured into contact with the virtual plane VP means bringing the surface 4a to be measured close to the virtual plane VP so that the surface profile can be measured. Since the virtual plane VP is virtual, it can enter the inside of the object to be measured, and such a state is also included in the measurable contact state.

This adjustment is performed by detecting the position using a position detection unit, which is a device for optically detecting the position of the object surface, so as to move the object surface to a position determined as a focus position in advance. The position detector optically detects the position of the object surface, for example using the technique of active autofocus. The optical system of the position detection unit is configured integrally with an optical system that implements the surface profile measurement method.

In this adjustment, first, with the plane-parallel substrate 40 held on the sample table 7, for example, the reference point PO is adjusted to be a point on the virtual plane VP and a point on the reference plane _40a . The position is set to the focus position by the position detection unit. After that, the plane-parallel substrate 40 is removed and the object 4 is held on the sample stage 7 . While confirming the position of the surface to be measured 4a with respect to the focus position by the position detection unit, the position adjustment device moves the sample table 7 so that the focus position, which is the reference point _PO , is positioned on the surface to be measured 4a. do. This position adjustment is performed by automatically or manually operating a position adjustment device based on the output of the position detection section.

Based on the output of the position detector, this position adjustment is performed by automatically or manually operating the integrated optical system of the position detector and the optical system for performing the surface profile measurement method as a position adjusting device. It is also possible to For example, since a large silicon wafer may be sucked and placed on the rotating mechanism, the optical system of the position detection unit and the surface shape measurement method are integrated rather than installing the position adjustment device on the bottom surface. It may be preferable to move the optical system that performs the

The movement of the sample table 7 is performed by parallel movement along the direction of the line connecting the two focal points P _Q and P _L . In addition, in order to be able to handle various objects to be measured, the sample stage 7 is provided with a three-point support mechanism or the like for adjusting the angle of the mounting surface of the object to be measured on the sample stage 7. , tilt angle adjustment may be performed. Based on the movement distance η due to parallel movement of the sample table 7 until the adjustment is completed and the known thickness d of the plane-parallel substrate 40, the thickness D of the object 4 is D=d−η (when D is thick) or D = d + η (when D is thin).

In the off-axis hologram acquisition step (#3), the surface to be measured 4a at the position of the virtual plane VP is obliquely illuminated with the spherical wave illumination light Q, and the data of the object light O, which is the reflected light, is obtained from the off-axis reference light R. is acquired by the image sensor 5 as an object beam off-axis hologram _IOR .

Similarly, when the object 4 and the sample stage 7 are not on the optical path of the spherical wave reference light L, the data of the inline spherical wave reference light L that obliquely passes through the virtual plane VP and is incident on the image sensor 5 is off-axis. Using the reference beam R, it is acquired by the image sensor 5 as a reference beam off-axis hologram _ILR . The data of these two types of off-axis holograms I _OR and I _LR cannot be acquired at the same time. Also, the irradiation conditions of the off-axis reference beam R must be kept the same when both data are acquired.

In the object light hologram generation step (#4), the light receiving surface 50 of the image sensor 5 is obtained from the data of the two types of off-axis holograms I _LR and I _OR by calculation processing using the fact that the spherical wave reference light L is spherical wave light. An object light hologram g representing a light wave of the object light O is generated with (z=0) as the hologram plane.

In the measurement object light hologram generation step (#5), the object light hologram _g on the light receiving surface 50 (z=0) is converted into a hologram at the position (z=z 0 ) on the virtual plane VP by light propagation calculation, and further, It is rotationally transformed according to the inclination angle α _O of the virtual plane VP with respect to the light receiving surface 50 . Thereby, a reconstructed object light hologram ^hV for measurement on the virtual plane VP (z= _z0 ) is generated.

In the reference point detection step (#6), the object light hologram g is propagated by light propagation conversion, the position where the object light O is condensed is detected near the reference light _condensing point PL, and the position is converted into a virtual A reference point S1 for shape measurement is set as a true mirror image point of the illumination light condensing point _PQ with respect to the plane VP. The positional information of the reference point S1 can be said to be information obtained by calibrating the positional information of the reference light _condensing point PL. By using the positional information of this reference point S1, highly accurate measurement of the surface to be measured becomes possible.

In the spherical wave optical hologram generation step (#7), a hologram of spherical wave light emitted from the reference point S1 for shape measurement is analytically generated as a spherical wave optical hologram ^sV on the virtual plane VP. The spherical wave optical hologram ^sV is a computer realization of a reference plane in a conventional physical reference substrate, such as the reference plane in a Fizeau interferometer.

In the surface profile measurement step (#8), the surface profile of the measured surface 4a of the object 4 is calculated using the surface distribution of the phase difference between the reconstructed object optical hologram ^hV and the spherical wave optical hologram ^sV . The planar distribution of the phase difference is obtained by dividing the reconstructed object beam hologram ^hV by the spherical wave optical hologram ^sV to produce the complex amplitude in-line hologram ^JV _OS for measurement with respect to the object beam O and the spherical wave optical hologram ^sV : It is obtained as the surface distribution of the phase of the complex amplitude inline hologram J ^V _OS .

(Details of optical system settings)
The sample stage 7 and the optical system shown in FIG. 2 are set as follows. A reference light condensing point PL, which is the condensing point of the inline spherical wave reference light _L , is arranged on the optical axis of the image sensor 5, and an illumination light, which is the condensing point of the spherical wave illumination light Q, is located off the optical axis. A light condensing point _PQ is arranged. These arrangement settings of the light sources P _Q and P _L and the image sensor 5 are fixed thereafter. The setting of the virtual plane VP is a process of fixing the positional information of the virtual virtual plane VP as substantial information of the position of the sample stage 7 by combining the sample stage 7 and the parallel plane substrate 40 . Two right-handed orthogonal coordinate systems are set in the figure. An xyz coordinate system whose origin is set at the center of the light receiving surface 50 of the image sensor 5 and an x'y'z' coordinate system set on the virtual plane VP. The optical axis of the image sensor 5 is the z-axis.

The virtual plane VP is a plane that vertically bisects a line segment connecting the reference light condensing point _PL and the illumination light condensing point _PQ . _A reference point PO indicating the position of the plane-parallel substrate 40 is set at the intersection of the virtual plane VP and the optical axis. The sample stage 7 is adjusted so that the reference surface 40a of the parallel plane substrate 40 is in contact with the virtual plane VP when the parallel plane substrate 40 is fixed to the sample stage 7 . Whether or not such adjustment has been achieved is determined using the holograms of the object beam O and the reference beam L, which are reflected beams. The sample table 7 is adjusted as follows.

A plane-parallel substrate 40 having a reference plane 40a is fixed to the sample table 7 and illuminated with spherical wave illumination light Q, and data of reflected light from the plane-parallel substrate 40 is obtained by using off-axis reference light R to obtain off-axis object light. Obtained as hologram I _OR . The data of the in-line spherical wave reference light L that passes through the virtual plane VP and is incident on the image sensor 5 without being interrupted in the optical path without the reference plane substrate 40 (and the sample table 7, the same applies hereinafter) is obtained. , is obtained as a reference beam off-axis hologram _ILR using the off-axis reference beam R. The in-plane distribution of the phase difference between the holograms I _OR and I _LR is obtained, and the sample stage 7 is adjusted by changing the position (including the inclination or posture) of the sample stage 7 so as to reduce the in-plane variation of the phase difference. I do. The plane distribution of the phase difference is obtained as the phase distribution of the complex amplitude in-line _{hologram JOL obtained by dividing the object light off-axis hologram IOR by the reference light off-axis hologram ILR .}

More specifically, the condensing points of the inline spherical wave reference light L and the spherical wave illumination light Q are arranged, and initially, the inline spherical wave reference light L and the off-axis reference light are arranged without the plane-parallel substrate 40. Record the interference fringes _{ILR produced} by R. Next, the plane-parallel substrate 40 is fixed to the sample stage 7 and illuminated with the spherical wave illumination light Q. As shown in FIG. The symmetrical point of the illumination light condensing point PQ with respect to the reference surface _40a of the plane-parallel substrate 40 approaches the reference light _condensing point PL. The position of the sample stage 7 (distance z _O and generally two-direction inclination α _O ) is adjusted so as to coincide with a plane that vertically bisects the line segment connecting the light point P _L and the illumination light condensing point P _Q . After mechanical adjustment, the interference fringes IOR formed by the object light O, which is the reflected light from the reference surface 40a, and the off-axis reference light _R are recorded.

Spatial frequency filtering is performed to extract from each interference fringe I _OR and I _LR the complex amplitude off-axis holograms J _OR and J _LR representing the respective real image components, and J _OR is divided by J _LR to obtain the complex amplitude in-line hologram J Get an _OL . The phase (θ _O −θ _L ) of the complex amplitude inline hologram J _OL represents the phase difference between the inline spherical wave reference beam L and the object beam O at the light receiving surface 50 . When the symmetrical point of the illumination light condensing point PQ approaches the reference light condensing point _{P L} , the planar distribution of the phase component exp[i(θ _O −θ _L )] of J _OL changes on the light receiving surface 50. It approaches a surface distribution with a small constant value. Further, the phase component exp[i(θ _O −θ _L )] has a surface distribution in which the value changes more drastically as the symmetrical point of the point P _Q is farther from the point P _L .

The distance between the symmetrical point of the point _PQ and the reference beam condensing point PL is the resolution δ ₌ λ/(2NA) or more in the direction perpendicular to the z-axis, or the depth of focus DOF=λ/(2NA ² ) above, the distribution of the phase component exp[i(θ _O −θ _L )] vibrates on the hologram plane as it moves away. Here, NA is the numerical aperture of the recorded hologram, and its numerical value is determined by the size of the aperture of the image sensor 5 and the distance from the image sensor 5 to the condensing point.

The distance z _O and the tilt angle α _O are adjusted so that the change in the phase component exp[i(θ _O −θ _L )] of the complex amplitude inline hologram J _OL is sufficiently small, and the reference surface 40 a of the plane-parallel substrate 40 is is determined as a virtual plane VP, and the adjustment of the sample table 7 is completed. The reference light L and the illumination light Q are symmetrical with respect to the adjusted reference surface 40a, and the x'y' plane distribution of the phase difference (θ _O −θ _L ) between the illumination light Q and the reference light L on the reference surface 40a is becomes a nearly constant value with small changes.

(Measurable shape)
The surface shape measuring method of the present embodiment uses a virtual plane as a reference plane based on the phase difference between the phase of the illumination light Q on the virtual plane VP and the object light O, which is the reflected light of the illumination light Q from the surface 4a to be measured. It is based on the principle that the height of the surface to be measured 4a from VP is obtained, and at the same time, the parallelism of the surface to be measured with respect to the bottom surface is measured on the basis of the bottom surface of the object to be measured. It is also possible to measure the thickness, inclination angle, and concave or convex shape of the object to be measured. In order to realize measurement based on this principle, the positional information of the reference point S1 for shape measurement, which is information obtained by correcting (calibrating) the positional information of the spherical wave reference light _L and the reference light condensing point PL in the computer. is used, and the position information of the reference point S0 for tilt angle and curvature measurement, which will be described later, is used.

As a means for setting one point on the surface 4a to be measured to a point on the virtual plane VP, the position detection section and the position adjustment device are used as described above. The position detection unit can detect the surface position on the surface to be measured 4a with an accuracy of several tens of nanometers or less using active autofocus technology, and the position adjustment device adjusts the position so that the set gap ε is several tens of nanometers or less. can be adjusted to suppress the influence of such a set interval ε.

When the position adjustment device is manual, for example, a moving table with a micrometer and the sample table 7 are integrated, and while observing the differential detection signal obtained by the focusing means of the position detection unit, the signal value becomes zero. A device that manually adjusts the moving table may be used. In the case of automatic adjustment, for example, as the electrically driven sample table 7, one driven by a step motor or one driven by a piezo actuator is used, and the differential detection signal obtained by the focus means of the position detection unit is driven. The signal may be used to electrically drive the sample stage 7 to perform a servo operation for adjustment.

(Hologram data and its processing)
Data processing will be explained based on mathematical expressions. The off-axis reference beam R, the in-line spherical wave reference beam L, and the object beam O associated with the hologram to be processed are represented by the following equation (1) ( 2) Represented by (3).

The light intensity I _OR of the combined light produced by the object light O and the off-axis reference light R, and the light intensity I _LR of the combined light produced by the in-line spherical wave reference light L and the off-axis reference light R are given by the following equation (4) ( 5). These light intensities I _OR and I _LR are obtained as hologram data through the image sensor 5 .

The above equations (4) and (5) are respectively Fourier-transformed into representations in the spatial frequency space, filtered by a band-pass filter to extract only the third term from the right side of the above equations, and subjected to inverse Fourier transformation. By this processing, an object beam complex amplitude hologram J _OR recording the object beam O and a complex amplitude hologram J _LR recording the inline spherical wave reference beam L are obtained as shown in the following equations (6) and (7).

A complex amplitude inline hologram _JOL of the object light O with respect to the inline spherical wave reference light L is obtained as shown in the following formula (8) by dividing the above formula (6) by the formula (7). This division process is a process of performing phase subtraction, that is, a process of performing frequency conversion, and is a process of heterodyne modulation.

The in-line spherical wave reference beam L is a reference beam for acquiring and storing the data of the reference beam R as a reference beam hologram _ILR , which is an off-axis hologram, and also serves as a reference beam in digital processing of hologram data. have When obtaining a plurality of object beam holograms I ^j _OR while maintaining the off-axis reference beam R under the same condition, it is _{sufficient to obtain a single off-axis hologram I LR .}

(Component of in-line spherical wave reference light L and multiplication factor)
Next, in equation (8), by multiplying both sides by a multiplication factor L ₀ (x, y) exp(iφ _L (x, y)), amplitude modulation by the amplitude factor L ₀ and phase factor exp(iφ _L ) is performed, and an object light hologram g(x, y) representing the light wave of the object light O on the hologram plane (z=0) is obtained as shown in the following equation (9), and the object light O is reproduced. be.

This multiplication process is a process of removing the component of the inline spherical wave reference light L from the above equation (8), and generates a hologram g containing only the light wave of the object light O. FIG. The term hologram is used to mean that it contains all the data necessary to reconstruct the light wave, and so on.

The above multiplication factor L ₀ exp(iφ _L ) is called the in-line reference beam hologram j _L . This is because the multiplication factor L ₀ exp(iφ _L ) is a spherical wave, and the spherical wave emitted from the condensing point P _L of the in-line spherical wave reference light L propagates in the air and reaches the image sensor 5 . because it represents This hologram _jL propagates through the air and reaches the image sensor 5 as a spherical wave. Therefore, the multiplication factor is obtained analytically using the positional information of the focal point _PL .

Note that when the in-line spherical wave reference light L passes through the beam coupler 3 or the like in a place other than air, as in the optical system in FIG. becomes. In this case, the hologram _jL is calculated by light propagation calculations using plane wave expansion (described below).

(Measurement of distance ρ, _zO and inclination angle _αO , plane wave expansion, calculation of light propagation)
The surface profile is measured by reproducing reflected light from the surface to be measured at a position parallel to the surface to be measured, that is, a virtual plane. The reproduction requires the distance z _O to the virtual plane VP, the inclination α _O of the virtual plane VP, and the distance ρ to the reference light condensing point P _L with respect to the light receiving surface 50 of the image sensor 5 . These values are determined with high accuracy by image recording and reconstruction of the target using holography.

In the optical system shown in FIG. 2, for example, a planar target comprising a pattern of precisely known dimensions formed by an opaque thin film on a transparent planar glass substrate is placed on the pre-adjusted sample stage 7 so that the pattern is in contact with the virtual plane VP. Next, the pattern side is directed toward the sample table 7 and fixed. Here, it is assumed that the sample stage 7 has an opening around the optical axis so that the in-line spherical wave reference light L can pass through. The off-axis object beam hologram _IOR is recorded with the beam transmitted through the target illuminated by the in-line spherical wave reference beam L as the object beam O. The object light hologram g on the light receiving surface 50 is obtained from the holograms I _OR and I _LR , and the light propagation calculation and rotational transformation of the object light g are performed as described below to reproduce a focused image at the target surface position.

By Fourier transforming the object light hologram g(x, y) of the above equation (9), the light wave of the object light O is expanded into plane waves, and the spatial frequency spectrum G(u, v) of the object light O is obtained by the following equation (10) ). The spatial frequency spectrum H(u,v) of the object light _O at the position z= _zO is obtained by the following equation (11) using G(u,v). If h(x, y, z ₀ ), h is obtained by the following formula (12).

The spatial frequency spectrum H ^V (u′, v′) after rotational transformation by the tilt angle α _O is expressed by the following equation (13), and the rotational transformation Jacobian J(u′, v′) is expressed by the following equation (14). Therefore, the reproduced object light h ^V (x', y', z _O ) after rotation conversion is given by the following equation (15).

The reproduced object beam ^hV includes the distance z ₀ , _ρ and the tilt angle α 0 as parameters. The distance _z0 is obtained from the z-coordinate value of the reproduction plane on which the focused reproduced image is obtained at least at the reference point _P0 , and the distance ρ is obtained from the parameter values when the dimensions of the focused reproduced image match the actual dimensions of the target. Also, the tilt angle α ₀ is obtained as the value of the rotation conversion angle when the in-focus reproduced image is obtained on the entire surface.

(Determination of reference points for shape measurement and highly accurate determination of virtual planes)
A highly accurate determination of the virtual plane VP based on hologram data will be described. First, let's talk about distance and measurement accuracy. The in-line spherical wave reference beam _L is used only for reproducing the hologram, and the distance ρ to the reference beam converging point PL is measured in mm. The shape measurement does not use the reference light condensing point PL, but rather uses a reference point S1 for shape measurement that is newly set by searching in the vicinity of the reference light condensing point _{P L and} a reference point light source set there. Use This reference point S1 is used as a true mirror image point of the illumination light condensing point _PQ . The reference point S1 is determined by post-processing on a computer using the following correlation function calculation so as to be a true mirror image point of the illumination light condensing point _PQ .

As shown in the flowchart of FIG. 3, in the propagation step (#61), the object light hologram g for the object light O, which is the reflected light of the illumination light Q, is converted to the position _z of the reference light condensing point PL by light propagation calculation. = ρ and let the generated hologram be the evaluation hologram h0 = h(x, y, ρ).

Next, in the in-plane condensing point position detection step (#62), the object light O (illumination light A condensing point where the reflected light of Q) is condensed is detected and set as a temporary condensing point P1 (x1, y1, ρ).

In the optical axis direction focal point position detection step (#63), the evaluation hologram h0=h(x, y, ρ) is propagated to a plurality of positions in the optical axis direction by light propagation calculation, and temporary light is condensed at each position. The correlation function calculation is performed by fixing the position (x1, y1) of the point P1 in the plane perpendicular to the optical axis. From the calculation result at each position in the optical axis direction, the condensing point of the object light O in the optical axis direction is detected, and the position z1=ρ+Δρ of the condensing point P1 in the optical axis direction is detected.

In the reference point setting step (#64), the detected condensing point P1 (x1, y1, z1) is set as the reference point S1 for shape measurement, and the reference point light source is set there.

The above processing will be described using mathematical expressions. The evaluation hologram h0=h(x, y, ρ) is given by the following equation (16) (#61). The probe function fp is a virtual point source fp=δ(x−x1)δ(y−y1) placed at coordinates (x1, y1, ρ). Correlation function C is given by the following equation (17).

Coordinates (x1, y1) are obtained with much higher accuracy than the resolution δ=λ/(2NA) as the parameter values that maximize the absolute value |C(x1, y1, ρ)| of the correlation function (#62). Next, the value of (x1, y1) is fixed in the above equation (16), ρ is changed to the parameter z1, and the value of z1 when the absolute value |C(x1, y1, z1)| becomes maximum is , the depth of focus DOF=λ/(2NA ² ), z1 is found, the position coordinates (x1, y1, z1) of the mirror image point P1 are obtained (#63), and the reference point S1 is set (# 64).

The spherical wave light generated by the point light source set at the reference point S1 is assumed to be an inline spherical wave reference light L1. The phase of the reference light L1 can be accurately calculated using the analytical solution of the spherical wave, and the change in the phase of the illumination light Q and the change in the phase of the reference light L1 can be accurately calculated over the entire virtual plane VP. matches each other.

The height distribution t(x', y') of the measured surface 4a measured with reference to the virtual plane VP is obtained from the optical path difference between the illumination light Q reflected by the measured surface 4a and the illumination light Q reflected by the virtual plane VP. sought. A position within a wavelength λ of light in space is represented by a phase value of the light. If the optical path difference is within the wavelength λ of light, the optical path difference can be represented by the phase difference of light. Information on the height distribution t is obtained by comparing the phase θ _O of the object light O, which is the reflected light of the illumination light Q from the measurement target surface 4a, and the phase θ _Q of the illumination light Q on the virtual plane VP. can be done. As the phase θ _Q of the illumination light Q, the phase θ _L1 on the virtual plane VP of the spherical wave emitted from the reference point S1 can be analytically derived and used.

The height distribution t is obtained as follows. A phase θ _O is obtained from the reconstructed object light hologram h ^V (x′, y′), and a phase θ _L1 is obtained from the spherical wave light hologram s ^V (x′, y′) of the reference light L1. The phase difference (θ _O −θ _L1 ) is obtained as the phase of the measurement hologram J ^V _OS (x′,y′)=h ^V /s ^V constructed by dividing h ^V by s ^V . The height distribution t is obtained by the following equation (18) using the phase difference (θ 0 _{-θ L1} ). Here, the angle β(x', y') is the incident angle of the illumination light Q at the coordinates (x', y').

(Position detection of surface to be measured)
Next, detection of the position of the measured surface 4a of the object 4 held on the sample stage 7 when setting the measured surface 4a to the position of the virtual plane VP will be further described. FIGS. 4A, 4B, and 4C show an example of detecting the position of the surface 4a to be measured using the light beam reflection type position detector 8. FIG. This position detection unit 8 uses, for example, a technique similar to the active autofocus technique in cameras and optical discs.

FIG. 4(a) shows a state in which the position of the sample table 7 is adjusted so that the reference surface 40a of the plane-parallel substrate 40 held on the sample table 7 and the virtual plane VP are aligned, and a predetermined , the position of the irradiation point Pa is detected by the position detection unit 8. FIG. The position of the irradiation point Pa is preferably the same as the position of the reference point _PO , but may be different.

The position detection section 8 includes a detection optical system 80 , a detection light source 81 , a photodetector 82 and a signal output section 83 . The detection light source 81 emits a light beam B0 for position detection that is reflected from the reference surface 40a or the surface to be measured 4a.

The detection optical system 80 includes a condensing lens 8a for focusing the light beam B0 emitted from the detection light source 81 on the irradiation point Pa and for receiving the reflected light from the irradiation point Pa and entering the photodetector 82. and an imaging lens 8b that allows the The detection optics 80 also includes a plane-parallel substrate 40 that provides a reference surface 40a and an object 4 that provides a surface to be measured 4a, which together define the optical path of the light beam B0.

The photodetector 82 has a plurality (two in this embodiment) of detection elements 82A and 82B, each of which is responsible for detecting the light beam B0 and outputs electrical signals d1 and d2 according to the detected light intensity. to output The detection elements 82A and 82B are elements configured by dividing the detection element of the photodetector 82 into two by a straight dividing line.

The detection optical system 80 is set so that when the light beam B0 is reflected at the position of the irradiation point Pa, the reflected light is incident on the dividing line. The detection optical system 80 is set so that the reflected light is incident on either side of the dividing line, depending on whether the position where the light beam B0 is reflected is in front of or behind the irradiation point Pa. . Therefore, for example, when the electrical signals d1 and d2 represent numerical values of light intensity, if the reflection position of the light beam B1 is at a point (irradiation point Pa) on the virtual plane VP, then d1=d2, otherwise d1<d2. or d1>d2.

The signal output unit 83 determines the reflection position of the light beam B1 and the irradiation point Pa on the virtual plane VP based on the electrical signals d1 and d2 output by the detection elements 82A and 82B of the photodetector 82 according to the received light intensity. and output an electric signal d3 representing reflection position information. For example, if d1, d2, and d3 represent the signal strength and d3=d1-d2, d3=0 in FIG. 4(a). The electric signal d3 is used as a control signal for adjusting the position of the sample stage 7. FIG.

FIG. 4(b) shows the case where the parallel plane substrate 40 is replaced with the object 4. FIG. In this example, the thickness of the object 4 is thinner than the thickness of the plane-parallel substrate 40, and the surface to be measured 4a is farther than the virtual plane VP. The light beam B0 after being reflected by the surface 4a to be measured is incident on the detection element 82A above the dividing line of the photodetector 82 through the imaging lens 8b. In this case, d1>d2 and d3>0. When the measured surface 4a of the object 4 is closer than the virtual plane VP, d3<0. When the sample stage 7 is moved to the position where the output signal d3 becomes 0, the point on the surface 4a to be measured of the object 4 comes to be included in the virtual plane VP.

FIG. 4(c) shows a case where the bottom surface of the object 4 and the surface to be measured 4a are not parallel, and the surface to be measured 4a is angled with respect to the virtual plane VP. Even if the surface to be measured 4a, which is the reflecting surface of the light beam B0, has an angle with respect to the virtual plane VP at the irradiation point Pa, the light beam B0 reflected at the irradiation point Pa is detected by the effect of the imaging lens 8b. Incident on the parting line of the device 82 . Therefore, the measurement target surface 4a can be positioned without being affected by the inclination of the measurement target surface 4a of the object 4. FIG.

(Second Embodiment: Surface Profile Measuring Apparatus)
A surface shape measuring apparatus 1 according to a second embodiment will be described with reference to FIGS. 5 and 6. FIG. The surface shape measuring device 1 is a device that measures the shape of the surface to be measured of the object 4 using holography. and an image reproducing unit 12 for reproducing an image on the surface to be measured from the hologram obtained.

The data acquisition unit 10 includes an image sensor 5 that converts light intensity into an electrical signal and outputs it as hologram data, a sample table 7, and a position that detects the position of the measured surface 4a of the object 4 held on the sample table 7. It includes a detector 8 , a position adjusting device 71 for moving the sample stage 7 , an optical system 2 for propagating each light, and a beam coupler 3 . The image sensor 5 is connected to a computer 11 as a controller and memory.

After the position of the sample table 7 is adjusted using the plane-parallel substrate 40, the object 4 is held as a bottom surface reference. is adjusted so that it is tangent to the virtual plane VP. 5 and 6 show the state in which the position adjustment has been completed.

The optical system 2 is an optical system for acquiring a hologram. Two optical systems for spherical wave illumination light Q and in-line spherical wave reference light L, and an off-axis reference light R are arranged on both sides of the virtual plane VP. and an optical system. The beam combiner 3 consists of a cube-shaped beam splitter arranged in front of the image sensor 5 and propagates each light together with the optical system 2 .

The spherical wave illumination light Q is light that illuminates the surface 4a to be measured of the object 4 and causes the image sensor 5 to record the object light O that is the reflected light. The optical path of the illumination light Q is provided with a lens 21 for condensing parallel light and a pinhole plate 22 having a pinhole at the condensing position. The pinhole is the condensing point PQ of the illumination light _Q and the position of the point light source of the spherical wave light Q. As shown in FIG.

Similar to the illumination light Q, the optical path of the inline spherical wave reference light L is provided with a lens 25 for condensing parallel light and a pinhole plate 26 . The pinhole of the pinhole plate 26 is the condensing point PL of the in-line spherical wave reference light _L and the position of the point light source of the spherical wave light L. FIG. The in-line spherical wave reference light L becomes in-line light when viewed from the light receiving surface 50 of the image sensor 5 with respect to the object light O composed of the reflected light of the illumination light Q (the two lights L and O are alternately generated light and not treated simultaneously).

The object beam O and the in-line spherical wave reference beam L pass through the beam combiner 3 and enter the image sensor 5 head-on. When viewed from the direction perpendicular to the light receiving surface 50 of the image sensor 5, the illumination light condensing point _PQ and the reference light _condensing point PL appear at the same optical position and are inline.

The off-axis reference beam R enters the beam combiner 3 from its side, is reflected by the internal reflecting mirror 30 , and enters the image sensor 5 . The optical path is provided with a small-diameter lens 23 for diameter expansion and a large-diameter lens 24 for collimation, and off-axis reference light R formed in a spherical wave shape is generated.

The optical system 2 is set so that the illumination light condensing point _PQ and the reference light _condensing point PL are mirror images of each other with respect to the virtual plane VP. The optical system 2 is arranged such that the illumination light Q obliquely illuminates the surface 4a to be measured, and the object light O, which is the reflected light thereof, is incident on the image sensor 5, and the inline spherical wave reference light L illuminates the virtual plane VP. Each light is propagated so that it passes obliquely and enters the image sensor 5 .

The beam combiner 3 multiplexes the object light O or the in-line spherical wave reference light L and the off-axis reference light R to enter the image sensor 5 . The beam combiner 3 is not limited to a cube-shaped beam splitter, and a beam combiner with another configuration may be used.

(Position adjustment of sample stage by position detector and position adjustment device)
The position detection unit 8 detects the position of a point on the surface 4a to be measured in an area including the front and back of the virtual plane VP in the optical axis direction of the image sensor 5, and outputs the information as reflection position information. It is a range finder and uses active autofocus technology. The position detection unit 8 irradiates light focused on a specific irradiation point Pa in space, receives the reflected light, and receives a signal that can determine whether the reflected position of the reflected light is in front of or far from the irradiation point Pa. to output In this embodiment, a reference point _PO is set as the irradiation point Pa.

The position detection section 8 includes a detection light source 81 , a photodetector 82 , a condenser lens 8 a and an imaging lens 8 b that constitute a detection optical system, and a signal output section 83 . The detection light source 81 emits a light beam B0 for position detection that is focused on the irradiation point Pa by the condenser lens 8a. The light beam passing through the condensing lens 8a irradiates the surface 4a to be measured and becomes reflected light. The photodetector 82 detects the light beam B0 reflected from the surface 4a to be measured through the imaging lens 8b.

The photodetector 82 and the signal output section 83 are equivalent to the photodetector 82 and the signal output section 83 described with reference to FIGS. The signal output unit 83 generates a shift signal d3 indicating the distance between the position where the light beam B0 is reflected and the virtual plane VP based on the electrical signals d1 and d2 output by the photodetector 82, and outputs the shift signal d3. The shift signal d3 is output to the position adjustment device 71 as reflection position information. The photodetector 82 may be an arbitrary one composed of at least two divided detection elements. The photodetector 82 has a plurality of detection elements, and the signal output section 83 outputs from each of the plurality of detection elements. A configuration may be adopted in which the difference signal of the signals is used as the deviation signal.

The position adjustment device 71 moves the sample stage 7 based on the reflected position information from the signal output section 83 so that the detected point is included in the virtual plane VP. In other words, the position adjusting device 71 moves the sample table 7 forward so that the reflection point of the light beam B0 comes to the irradiation point Pa so that the irradiation point Pa is imaged on the photodetector 82 by the imaging lens 8b. or move backwards. As a means for moving the sample stage 7 using an electric driving device, for example, a piezo driving device that moves the sample table 7 with nanometer precision by pulse driving may be used.

The detection light source 81, the condenser lens 8a, the imaging lens 8b, and the photodetector 82, which are the optical system of the position detection unit 8, are structurally arranged together with the hologram acquisition optical system 2, the image sensor 5, and the beam combiner 3. are integrated and arranged on a common substrate. Considering the detection accuracy, it is desirable that the diameter of the light beam condensed by the condensing lens 8a and incident on the irradiation point Pa on the virtual plane VP be, for example, a spot of about 10 to 200 μm. Considering the detection sensitivity, the angle of incidence on the irradiation point Pa is desirably in the range of 20 to 60°, for example.

When the position of the sample table 7 is adjusted so that the surface 4a to be measured of the object 4, at least the surface 4a to be measured around the irradiation point Pa, is in contact with the virtual plane VP, the surface shape measuring method described in the first embodiment is performed. , the recorded hologram _IOR of the object light O, which is the reflected light, is obtained with the object 4 placed. Also, with the object 4 and the sample table 7 not on the optical path, the in-line reference beam L is used to obtain the recorded hologram _ILR of the off-axis reference beam R. FIG. The recording hologram _ILR of the reference beam _R may be obtained before the position of the sample stage 7 is adjusted, or may be obtained after the hologram IOR of the object beam O is obtained.

The image reproduction section 12 is provided in the computer 11 together with the data storage section 6 . The image reproduction unit 12 is configured with a software group and a memory for executing the surface shape measuring method described in the first embodiment. When the data acquisition unit 10 acquires the off-axis holograms I _OR and I _LR , they are processed by the surface profile measurement method described in the first embodiment to obtain surface profile measurement values.

Since the surface shape measuring apparatus 1 of this embodiment includes a cube-shaped beam coupler 3, the light propagation calculation of the light passing through the beam coupler 3 is performed by the plane wave expansion method in consideration of the refractive index of the beam coupler 3. need to do Processing related to the beam combiner 3 will be described below.

(Calculation of spherical wave after passing beam combiner)
In order to generate the object light hologram g from the complex amplitude hologram _JOL on the light receiving surface 50 of the image sensor 5, the light wave of the inline spherical reference light L (inline A reference beam hologram j _L ) is required. The in-line reference beam hologram j _L is not a spherical wave due to passing through the beam combiner 3 . Therefore, the light propagation calculation of the light wave from the position of the light converging point PL of the inline spherical wave reference light _L to the light receiving surface 50 which is the incident surface of the image sensor 5 is performed, and the inline reference, which is not the spherical wave, on the light receiving surface 50 is calculated. Generate light _L , the in-line reference light hologram jL.

Light propagation calculations are performed using plane wave expansion. Plane wave expansion of the reference light L at the condensing point P _L is propagated through the air and the beam combiner 3 to calculate each plane wave component on the light receiving surface 50 , and the calculated plane wave components are added to form an inline reference light hologram j Generate _L. A point light source b ₀ δ(x)δ(y) of the inline spherical wave reference light L exists in the xy plane at the position z=ρ of the condensing point P _L . The spatial frequency spectrum B(u,v) of this point source is a constant value b ₀ , ie B(u,v)=b ₀ . Therefore, due to the propagation of the plane wave, the inline spherical wave reference beam L at the light receiving surface 50 at z=0, ie, the inline reference beam hologram _jL , is given by the following equation (19).

n in (n/λ) in the above equation is the refractive index of the beam combiner 3 . The above equation (19) is a function of the distance ρ from the origin z ₌ 0 to the converging point PL and the dimension A of the beam combiner 3, but is independent of the distance from the origin to the beam combiner 3. In other words, the equation is the same no matter where the beam combiner 3 is placed. The above formula (19) is a theoretical calculation formula, and in actual calculation, it is necessary to perform light propagation calculation with the number of calculation points that satisfies the sampling theorem.

(Object light g(x, y) on the hologram plane)
The inline reference light hologram j _L of the above equation (19) obtained by the above procedure is the light wave of the inline spherical reference light L that has passed through the beam combiner 3 and reached the light receiving surface 50 . By multiplying the above formula (8) by the multiplication factor j _L =L ₀ (x, y) exp(i(φ _L (x, y)) consisting of this hologram j _L , the surface of the image sensor 5 (hologram surface, An object beam hologram g(x,y) representing a light wave of the object beam O in the xy plane, or plane z=0, is obtained similarly to equation (9) above: g(x,y)=J _OL × _jL .

(light propagation calculation)
A spatial frequency spectrum G(u, v) for the object light O is obtained as shown in the following equation (20) as a result of plane wave expansion for Fourier transforming the object light hologram g(x, y) on the hologram plane. In terms of expression, it becomes the same as the above formula (10). According to the light propagation calculation of the plane wave, the object light h(x, y) in the plane parallel to the light receiving surface 50 of the image sensor 5 at the position z= _z0 of the surface 4a to be measured of the object 4 is obtained by the following equation (21). can get.

u and v in the above equation (20) are the Fourier spatial frequencies in the x and y directions, respectively. The Fourier spatial frequencies w and _wn in the z direction can be obtained from the plane wave dispersion formula (relational expression between wave number and wavelength) as in the lower formula of the above formula (19). The dispersion equation includes the refractive index n in the form (n/λ) ² . The above formulas (20) and (21) are calculation formulas in consideration of the size A and the refractive index n of the beam combiner 3 existing on the optical path.

Since the object light h(x, y) parallel to the light receiving surface 50 of the image sensor 5 at the position z= _z0 of the surface 4a to be measured of the object 4 is obtained by the above equation (21), the above equation (13) ) to (18), high-precision determination of the virtual plane using the correlation function, and calculation of the height distribution, it is possible to perform surface profile measurement and obtain measurement results. The processing of the above equations (13) to (18) is processing of events in air, and it is not necessary to consider the influence of the refractive index n of the beam combiner 3 and the like.

(Third embodiment)
A surface shape measuring apparatus 1 according to a third embodiment will be described with reference to FIG. The surface shape measuring apparatus 1 of this embodiment is the same as that of the second embodiment except for the configuration of the position detection unit 8 that detects the position of the surface 4a to be measured.

The position detection unit 8 includes a detection light source 81 for measuring the time-of-flight (TOF) of light irradiated on the surface 4a to be measured and detecting the position of the surface 4a to be measured, and two light detectors for detecting light beams. A device 82 (82A, 82B), a beam splitter 8c, and a signal output section 83 are provided. The detection light source 81 emits a modulated light beam used for detecting the position of the surface 4a to be measured. A pulse-modulated laser beam, for example, is used as the light beam.

A light beam emitted from the detection light source 81 is split into two light beams by the beam splitter 8c. One of the light beams that has been turned and separated by the beam splitter 8c is incident on one of the photodetectors 82A. The other light beam that has been split by going straight through the beam splitter 8c propagates toward the surface to be measured 4a. incident on the photodetector 82B. This other light beam reciprocates on the same optical path between the beam splitter 8c and the surface to be measured 4a. In the round trip optical path, the optical path is bent by the internal reflecting mirror 30 of the beam combiner 3, so that the light is incident on the surface to be measured 4a and reflected from the surface to be measured 4a so as to be substantially perpendicular to the virtual plane VP. The detection optical system of the position detection unit 8 is set so that the light beam directed toward the surface 4a to be measured passes through the irradiation point Pa on the virtual plane VP.

Photodetectors 82A and 82B detect the separated light beams and output signals d4 and d5, respectively. Signals d4 and d5 are, for example, the number of pulse counts from the emission of the light beam by the detection light source 81 to the reception of each separated light. The positional information of the reflection point on the surface 4a to be measured is obtained from the difference in the time (number of pulse counts) during which the light emitted from the detection light source 81 is received by the two photodetectors 82A and 82B. The signal output unit 83 outputs previously acquired positional information of the reference surface 40a of the plane-parallel substrate 40 when the reference surface 40a is set to the position of the virtual plane VP, and thus the positional information of the irradiation point Pa on the virtual plane VP. , the time difference between the signals d4 and d5 is stored in the memory.

The signal output unit 83 indicates the distance between the position of the surface to be measured 4a and the position of the virtual plane VP based on the positional information of the surface to be measured 4a and the positional information of the virtual plane VP stored in advance. Generate a shift signal d3. The shift signal d3 is output to the position adjusting device 71 as reflected position information.

The position adjusting device 71 moves the position of the sample table 7 forward or backward so that the shift signal d3 becomes zero. When the shift signal d3 becomes zero, the surface 4a to be measured of the object 4, at least the surface 4a to be measured around the irradiation point Pa, comes into contact with the virtual plane VP, and the adjustment of the position of the sample stage 7 is completed.

The detection light source 81, the photodetector 82 (82A, 82B), and the beam splitter 8c, which are the optical system of the position detection unit 8, are structurally It is integrated and arranged on a common substrate. The time difference obtained from the signals d4 and d5 may be a femtosecond time difference when measuring distance with an accuracy of about 100 nm, for example.

(Modification)
A surface shape measuring apparatus 1 according to a modification of the third embodiment will be described with reference to FIG. This modification differs from the third embodiment using modulated laser light in that the detection light source 81 of the position detection unit 8 emits white light. The detection optical system of the position detection unit 8 is the same as in the third embodiment in that the light beam directed toward the surface 4a to be measured is set so as to perpendicularly pass through the irradiation point Pa on the virtual plane VP. The position detection section 8 includes a detection light source 81 , a photodetector 82 , a beam splitter 8 c , a reflecting mirror 8 e and a signal output section 83 .

A white light beam emitted from the detection light source 81 is split into two light beams by the beam splitter 8c. One of the light beams, which has been turned and separated by the beam splitter 8c, is vertically reflected by the reflecting mirror 8e, passes through the beam splitter 8c, and enters the photodetector . The other light beam that has been split by going straight through the beam splitter 8c propagates toward the surface to be measured 4a, and after the reflected light reflected by the surface to be measured 4a returns to the beam splitter 8c, the optical path is bent to become light. Incident on detector 82 .

The detection optical system of the position detection unit 8 is arranged so that the total optical path length of the light beam reflected by the reflecting mirror 8e is equal to the total optical path length of the light beam reflected at the irradiation point Pa on the virtual plane VP. is set. Therefore, if the other light beam is reflected not at the irradiation point Pa but at a point on the surface to be measured 4a that is displaced from that position, the magnitude of the shift is the total optical path length of the two light beams. measured as the difference between

The photodetector 82 acquires data d6 relating to the interference between the white lights of the two light beams, and outputs the data d6 to the signal output section 83. FIG. Based on the data d6, the signal output section 83 generates a deviation signal d3 indicating the distance between the position of the surface 4a to be measured and the position of the virtual plane VP. The shift signal d3 is output to the position adjusting device 71 as reflected position information.

The detection light source 81, the photodetector 82, the beam splitter 8c, and the reflecting mirror 8e, which are the optical system of the position detection unit 8, are structurally common with the hologram acquisition optical system 2, the image sensor 5, and the beam combiner 3. are arranged integrally on the substrate of the The signal processing configuration of the photodetector 82 and the signal output section 83 can be arbitrarily changed, and the signal output section 83 may be incorporated in the photodetector 82 .

(Fourth embodiment)
A surface shape measuring apparatus 1 according to a fourth embodiment will be described with reference to FIGS. 9 and 10. FIG. In the surface shape measuring apparatus 1 of the present embodiment, the optical system 2 for acquiring the hologram in the surface shape measuring apparatus 1 of the second embodiment is replaced with a condensing lens 27 for condensing the object light O and the inline spherical wave reference light L. The optical system 2 includes a pupil plate 27a arranged at a light condensing position by the condenser lens 27 to limit the amount of light passing therethrough, and an imaging lens 27b arranged in combination with the pupil plate 27a. The two lenses provided in front and behind the pupil plate 27a are lenses for forming an image of the object light O and the inline spherical wave reference light L on the image sensor 5. FIG.

Moreover, the surface shape measuring apparatus 1 of the present embodiment includes the position detection section 8 and the position adjustment device 71 similar to those of the second embodiment. The detection light source 81, the condenser lens 8a, the imaging lens 8b, and the photodetector 82, which are the optical system of the position detection unit 8, are structurally arranged together with the hologram acquisition optical system 2, the image sensor 5, and the beam combiner 3. are integrated and arranged on a common substrate.

If a hologram with a large aperture can be recorded, it will be possible to measure the surface shape of a large object. If the reflected light is focused using a lens as in this embodiment, a single image sensor 5 can record a large-diameter hologram. A condensing lens is used to alternately project the in-line spherical wave reference beam L and the object beam O onto the light-receiving surface of the image sensor 5, and the interference fringes produced by the off-axis reference beam R are respectively recorded. The width of the spatial frequency band of the recorded hologram can be adjusted by opening and closing the pupil of the pupil plate 27a. When the surface to be measured is smooth and highly flat, the spatial frequency bandwidth is narrow, and when the surface to be measured has minute unevenness, the bandwidth is wide. The pupil plate 2 may be adjusted according to the condition of the surface 4a to be measured and the purpose of measurement.

Since the two lenses, the condenser lens 27 and the imaging lens 27b, form an image of the light on the surface to be measured on the light receiving surface of the image sensor 5, the shape of the surface to be measured can be obtained without reproducing the object light. Observation and shape measurement become possible.

(Fifth embodiment)
A surface shape measuring apparatus 1 according to a fifth embodiment will be described with reference to FIG. The surface shape measuring apparatus 1 of this embodiment has a concave mirror 28 instead of the condenser lens 27, the pupil plate 27a, and the imaging lens 27b of the optical system 2 in the surface shape measuring apparatus 1 of the fourth embodiment. , a pupil plate 28a and an imaging lens 28b. Concave mirror 28 is, for example, a condensing ellipsoidal mirror.

The detection light source 81, the condenser lens 8a, the imaging lens 8b, and the photodetector 82, which are the optical system of the position detection unit 8, are structurally arranged together with the hologram acquisition optical system 2, the image sensor 5, and the beam combiner 3. are integrated and arranged on a common substrate.

In the present surface shape measuring apparatus 1 as well, the concave mirror 28 and the imaging lens 28b form an image of the object light O and the inline spherical wave reference light L on the image sensor 5 . In addition, a hologram with a large aperture can be recorded with a small image sensor, and the shape of the surface to be measured can be observed and measured without reproducing the object beam.

(Sixth embodiment)
A surface shape measuring apparatus 1 and a surface shape measuring method according to the sixth embodiment will be described with reference to FIG. The apparatus and method of this embodiment extend the range of heights that can be measured, and use light of different wavelengths (λ _j , j=1, 2) to achieve the extension. The optical system 2 for hologram acquisition of the surface shape measuring apparatus 1 of this embodiment is the same as the optical system 2 of the third embodiment or its modifications (FIGS. 7 and 8), but the beam combiner 3 and the image sensor 5, and two pairs of such wavelength filters and image sensors are provided. With this configuration, mixed light of two wavelengths can be used.

In addition, the device 1 of this embodiment includes a position detection section 8 and a position adjustment device 71 similar to those of the third embodiment or its modification. The detection light source 81, the photodetector 82, the beam splitter 8c, and the reflecting mirror 8e, which are the optical system of the position detection unit 8, are structurally common with the hologram acquisition optical system 2, the image sensor 5, and the beam combiner 3. are arranged integrally on the substrate of the Note that the position detection unit 8 may be the same as the position adjustment device 8 in the second embodiment.

A set of the wavelength filter F1 and the image sensor 51, which passes _one of the two different wavelengths λ1, is arranged on the side of the beam combiner 3 facing the incident surface 31 of the object light O. FIG. _A set of the wavelength filter F2 and the image sensor 52, which transmits the other wavelength λ2, is arranged on the side of the beam combiner 3 opposite to the surface on which the off-axis reference beam R is incident.

(Surface profile measurement using phase difference between light waves with different wavelengths)
In the surface shape measuring method of this embodiment, after the placement of the object 4 and the adjustment of the sample stage 7 are performed, the following processing is performed. Mixed light or individual light with different wavelengths λ _j , j=1, 2 allows the data of the object beam O and the inline spherical wave reference beam L to be converted into two types of off-axis holograms I ^j _OR for each wavelength λ ₁ , λ ₂ . , I ^j _LR , j=1,2. Next, for each wavelength λ ₁ , λ ₂ , a measurement hologram J _j ^V _OS =h _j ^V /s _j ^V , j=1, 2 is generated, and the two generated measurement holograms J _j ^V _OS , j=1,2 is performed. As a result of heterodyne conversion, a modulated wave HW=J ₁ ^V _OS /J ₂ ^V _OS is generated. Using the modulated wavelength λ _B =λ ₁ λ ₂ /(λ ₂ −λ ₁ ) and the modulated phase distribution θ _B (x′, y′)=θ ₁ −θ ₂ included in this modulated wave HW, A height distribution on the measurement surface 4a is obtained. Note that the modulated wave HW and the modulated wavelength λ _B are also called a composite wave HW and a composite wavelength λ _B , respectively.

The background and effect of the above processing will be described. For example, in the surface shape measurement using the phase of monochromatic laser light shown in the third embodiment, it is difficult to measure a height that greatly exceeds the light wavelength λ. In addition, for a step exceeding λ/2, an uncertainty of an integral multiple of λ/2 occurs in the height measurement value. By the way, a wave having a long wavelength can be generated by performing arithmetic processing on two light waves having different light wavelengths and having the same propagation direction. Using this wave phase can greatly expand the range of heights that can be measured.

Spherical wave illumination light Q with light wavelengths λ ₁ and λ ₂ emitted from the same point light source has the same light propagation direction at all points in space, and the phase components are exp(2πr/λ ₁ -θ ₁ ) and It is expressed as exp(2πr/λ ₂ -θ ₂ ). A wave having a phase component of exp(2πr/λ _B −θ _B ) can be created by dividing the spherical wave illumination light Q of light wavelength λ ₁ by the spherical wave illumination light Q of light wavelength λ ₂ . Here, λ _B and θ _B are given by the following equation (22). The wavelength λ _B coincides with the wavelength of the beat waves produced by the two illumination lights.

When the surface to be measured 4a is illuminated with two spherical waves having different wavelengths from the same light source position, the propagation directions of the two reflected lights emitted from each point on the measurement surface are the same. In addition, the propagation directions of two reflected lights emitted from microscopic surfaces on the measurement surface where interference and diffraction of light near the surface can be ignored also coincide. Therefore, dividing the reflected light of optical wavelength λ ₁ by the reflected light of optical wavelength λ ₂ produces a light wave of wavelength λ _B that functions in the same manner as illumination light Q, but with an increased wavelength. This indicates that surface shape measurement can be performed according to the measurement method shown in each embodiment by using the created wave of wavelength λ _B . If the phase difference between the phases θ _BO and θ _BL1 of the two waves of wavelength λ _B on the measured surface 4a and the virtual plane VP is expressed as Δθ _B (x′, y′)=θ _BO −θ _BL1 , the measured surface is given by the following equation (23). This equation (23) is equivalent to equation (18) for a single wavelength.

The above equation (23) is basically equivalent to equation (18) for a single wavelength. The surface shape measuring apparatus 1 and the surface shape measuring method of the present embodiment can arbitrarily decide during post-processing whether to use both data of holograms recorded for two wavelengths or to use either one of them. When using data of both wavelengths, formula (23) may be used, and when using data of a single wavelength, formula (18) may be used.

Holograms of different wavelengths can be one-shot recorded using the optical system of FIG. 12 and mixed light. In this case, an off _- axis reference beam _R2 for the optical wavelength λ2 is prepared in addition to the off _- axis reference beam _R1 for the optical wavelength λ1. In this optical system, in order to separate the _respective optical wavelength components, there _{are a} wavelength filter F1 that transmits light of optical wavelength λ1 and blocks light of optical wavelength λ2, and a wavelength filter F1 that transmits light of optical wavelength λ2 and blocks light of optical wavelength λ2. _A wavelength filter F2 that cuts off light of wavelength λ1 is used.

As another optical system 2 for the measurement method of this embodiment, for example, using the optical system of FIG. I ^j _OR and I ^j _LR may be obtained at different times for each wavelength with individual lights that are not mixed lights.

As another optical system 2, in the optical system of FIG. 7, an optical system for off-axis reference light R may be provided for each wavelength. In this case, the two off-axis reference beams R1 and R2 can be placed off-axis with each other, and holograms with different wavelengths can be one-shot recorded by the mixed beam. Separation into holograms by wavelength can be done by post-processing due to the effect of off-axis placement. A complex amplitude component of light wavelength λ ₁ and a complex amplitude component of light wavelength λ ₂ can be separately extracted from a one-shot recorded hologram by filtering in the spatial frequency domain.

It should be noted that if two optical systems arranged off-axis relative to each other are used for the off-axis reference beam R, the recordable measurement plane is narrower than when the optical system of FIG. 12 is used. Conversely, in the case of the optical system of FIG. 12, although the recordable measurement surface can be enlarged, the two holograms are recorded by separate image sensors 51 and 52, respectively. You will need to adjust the position of the light.

According to the surface shape measuring apparatus 1 and the surface shape measuring method of this embodiment, the synthesized wavelength λ _B =(λ ₁ λ ₂ )/(λ ₂ −λ ₁ ) is any of the original wavelengths λ ₁ , Since it is longer _than λ2, the measurable height range can be extended. The surface shape measuring apparatus 1 and the surface shape measuring method using light of different wavelengths are not limited to two wavelengths, and can be extended to devices and methods using three or more wavelengths. This method can perform measurement by post-processing the recorded hologram data, which is very different from the conventional method using beat waves. Therefore, for example, in the case of three wavelengths λ ₁ , λ ₂ , λ ₃ , two wavelengths are selected by post-processing, and a plurality of combinations such as the difference (1/λ ₁ -1/λ ₂ ) are created, or three wavelengths can be used to create a plurality of combinations such as sum and difference (1/λ ₁ +1/λ ₂ -1/λ ₃ ), and measurement can be performed by interpolating the measurement regions with each combination. can.

(Seventh embodiment)
A surface shape measuring apparatus 1 according to a seventh embodiment will be described with reference to FIG. Since the surface shape measuring device 1 of the present embodiment can be embodied by, for example, the surface shape measuring device 1 shown in FIGS. 5 to 12, these figures are also referred to. The surface shape measuring apparatus 1 includes a data acquisition unit 10 that acquires the hologram of the surface 4a to be measured, and an image reconstruction unit 12 that reconstructs an image of the surface 4a to be measured from the hologram acquired by the data acquisition unit 10. there is The surface shape measuring apparatus 1 further includes a control unit 11 comprising a computer that controls the data acquisition unit 10 and the image reproduction unit 12, and a memory 11a that stores calculation programs such as FFT, control data, and the like. .

The data acquisition unit 10 includes an optical system 2 that generates and propagates light, a beam combiner 3 that is a cube beam splitter used as a beam combiner, and an image that converts the light intensity into an electrical signal and outputs it as hologram data. It has a sensor 5 and a data storage unit 6 for storing data acquired by the image sensor 5 . The data storage unit 6 is provided in the control unit 11 together with the image reproduction unit 12 . The beam coupler 3, together with the optical system 2, propagates each light for hologram acquisition.

The data acquisition unit 10 also includes a sample stage 7 whose position and attitude can be adjusted in relation to the arrangement configuration of the optical system 2 and the image sensor 5, a position adjustment device 71 that adjusts the position of the sample stage 7, a parallel A position detection unit 8 detects the position of the reference surface 40a of the flat substrate 40 and the position of the surface 4a to be measured of the object 4 and outputs the result to the position adjustment device 71 as a signal for position adjustment.

The image reproduction unit 12 has hologram generation units 13 to 16 and 18, a reference point detection unit 17, a shape measurement unit 19, and a display unit 20 for performing the processes shown in FIGS. is doing.

The complex amplitude hologram generator 13 removes the component of the off-axis reference beam _R from the object beam off-axis hologram IOR and the reference beam off-axis hologram _ILR to obtain the complex amplitude of the object beam O and the inline spherical wave reference beam L. Generate an in-line hologram _JOL .

Based on the fact that the light emitted from the reference light condensing point PL is a spherical wave, the computational reference light hologram generator 14 generates a light wave of the inline spherical wave reference light _L on the light receiving surface 50 which is the light receiving surface of the image sensor. Create an in-line reference beam hologram j _L representing

The object light hologram generator 15 uses the inline reference light hologram _jL to generate an object light hologram g representing the light wave of the object light O on the light receiving surface 50 from the complex amplitude inline hologram _JOL .

The reproduced object light hologram generator 16 converts the object light hologram g into a hologram at the position of the virtual plane VP by light propagation calculation, and converts the converted hologram to a virtual plane tilt angle α _O to generate a reconstructed object beam hologram ^hV for measurement in the virtual plane VP.

The reference point detection unit 17 performs light propagation calculation of the object light hologram g, detects the condensing point of the object light by correlation function calculation, and sets the point as the reference point S1 for shape measurement.

The analysis light hologram generator 18 analytically generates a spherical wave light hologram ^sV which is a hologram on the virtual plane VP of the spherical wave corresponding to the inline spherical wave reference light L emitted from the reference point S1.

The shape measurement unit 19 divides the reproduced object light hologram ^hV by the spherical wave light hologram ^sV to generate a measurement hologram ^JVOS relating to the object light O and the spherical wave light hologram ^sV , and _calculates the complex amplitude for measurement. The height distribution on the measured surface 4a of the object 4 is obtained from the phase distribution of the in-line hologram J ^V _OS . Further, the shape measuring unit 19 calculates the thickness D of the object 4 obtained from the moving distance η of the sample table 7 and the known thickness d of the plane-parallel substrate 40, and the height distribution on the surface 4a to be measured. are used to calculate various quantities related to the shape of the object 4 . For example, when the object 4 is a silicon wafer, the various quantities include the thickness distribution of the surface to be measured with respect to the bottom surface of the wafer, the parallelism evaluation value, and the like. Measurement points on the surface to be measured can be specified at desired intervals to obtain a large number of measured values, and the shape measuring section 19 calculates a large number of measured values and various statistical quantities.

The display unit 20 displays an image obtained by the image sensor 5, an intensity image of each hologram, a phase distribution image, and the like. The data of the object beam off-axis hologram _IOR and the reference beam off-axis hologram _ILR stored in the data storage unit 6 are processed by the image reproducing unit 12 and displayed on the display unit 20 . The display unit 20 is an FPD such as a liquid crystal display device, displays data other than images, and serves as a user interface. Each unit of the image reproduction unit 12, except for the display unit 20, is configured using software including a program operating on a computer and its subroutine group.

(Features and application examples of the present invention)
In the surface shape measuring apparatus and method according to the present invention, points on the surface to be measured 4a are detected using the position detection unit and the position adjustment apparatus, and the detected points are included in the virtual plane (VP). The position of the sample stage is adjusted to adjust the position of the object in real space, and the data of the object beam off-axis holograms I _OR and I ^j _OR are acquired. In addition, after obtaining the necessary hologram data, processing is performed in a non-real space on the computer, that is, by post-processing by calculation processing performed in the computer, processing to exceed the limit of measurement accuracy in the real space, Reconstruction of an object image and surface shape measurement are realized with high accuracy.

In the post-processing of hologram data by a computer, various methods are used to determine whether the processing is appropriate and whether necessary processing has been completed. For example, regarding the data after processing, the clarity (degree of focusing) of the reproduced image, whether or not the reproduced image reproduces known dimensions, the maximum or minimum value of the correlation function between data, how the phase distribution appears , how the fringe pattern appears in spatial frequency space, and so on.

How the interference fringe pattern appears, for example, is the point detected by the position detection unit (for example, the point corresponding to the reference point _PO ) or the reference point _PO itself on the virtual plane VP as a fixed point of rotation of the object light hologram. When data is rotated in a three-dimensional space and the object light hologram is a hologram on a predetermined plane, for example, a virtual plane VP, this is used to determine whether the rotation is appropriate. In this case, it can be judged, for example, by a change in appearance such as an increase in the distance between interference fringes between the object light and the spherical wave light on the virtual plane VP.

Further, in order to determine whether the peripheral area of the point detected by the position detection section on the surface 4a to be measured of the object 4 is concave or convex, the object light hologram is rotated or moved along the optical axis. Therefore, the appearance of the phase distribution, the appearance of the interference fringe pattern, and the like can be used. For example, the data of the object light hologram is moved in the three-dimensional space in the direction of the optical axis, and at the point of interest (for example, the intersection of the optical axis and the surface to be measured 4a), the object light and the spherical wave light on the virtual plane VP during movement. By using the information on unevenness, such as whether the interference fringes of 1. are coming out or being absorbed, the measurement range of the surface to be measured can be expanded by data processing even in the case of measurement with single-wavelength light.

As an application example according to the implementation of the present invention, for example, when the main surface (surface to be measured) of a substantially circular silicon semiconductor substrate is deformed into a convex or concave shape, in addition to determining whether the deformation is convex or concave, and measure the radius of curvature of the deformed curved surface. As a preliminary step, a plane-parallel substrate having a reference plane with a predetermined flatness is used as the object 4, the sample table is adjusted so that the reference plane is in contact with the virtual plane VP, and a spherical wave from the surface to be measured composed of the reference plane is Data of the object light OP, which is the reflected light of the illumination light ^Q , is acquired as an object light off _- axis hologram I _ORP using an image sensor. An object light hologram ^gP representing the light wave of the object light _OP is generated from the data of this object light off _- axis hologram IORP and the already acquired reference light off _- axis hologram _ILR , and the object light hologram _gP is calculated by computational processing. is optically propagated and rotationally transformed to generate a reconstructed object light _{hologram hPV} on the virtual plane VP, and by computational processing, the object light hologram _g0 is optically propagated and converted to the position where the object light ^OP is condensed. is detected and its position is regarded as the mirror image point of the illumination light condensing point _PQ with respect to the virtual plane VP, and this condensing point is defined as the reference point S0 for tilt angle and curvature measurement. This reference point S0 is detected using a correlation function in the same manner as the detection of the reference point S1 for shape measurement (see FIG. 3).

When the information of the reference point S1 for shape measurement is obtained, the fine height distribution of the surface to be measured can be calculated. Ψ (the first-order term when the height distribution function of the surface to be measured is obtained as the height distribution with respect to the reference surface and the height distribution function is Taylor-expanded) and the radius of curvature R _S ( Information of the secondary term when the above height distribution function is Taylor-expanded cannot be obtained. This is because when the reference point S1 for shape measurement is detected using a correlation function, the position of the peak of the correlation function may be in any direction in the three-dimensional space, depending on the tilt angle of the substrate and the shape of the convex or concave surface. because it moves. Information on the inclination angle Ψ of the substrate and information on the radius of curvature R _S of the substrate relating to the shape of the convex or concave surface of the substrate are lost from the information of the reference point S1 for shape measurement. The height information of the third or higher order term of is left, and the following root mean square (RMS) can be obtained.

If you look at the standard specifications for dimensional standards for silicon wafers (standard specifications for dimensional standards for silicon mirror wafers, JEITA EM-3602), you can find not only the root mean square root value RMS of the height of the surface to be measured relative to the reference surface, The radius of curvature of the surface to be measured and the angle of inclination with respect to the reference plane are required. Consider the main surface of a semiconductor substrate as the surface 4 a to be measured of the object 4 . Data of object light, which is reflected light from the main surface of the semiconductor substrate, is acquired, and the reference point S1 for shape measurement of the main surface of the semiconductor substrate is obtained using the method shown in FIG. Compare with the reference point S0 for curvature measurement. From the position coordinates S0 (x0, y0, z0), S1 (x1, y1, z1) of the two reference points, the distance L _xy =√((x1- x0) ² +(y1-y0) ² ), and the distance L _z =z1-z0 in the direction of the optical axis z (see FIG. 2 for xyz coordinates). √(*) represents the square root of (*).

A plane that vertically bisects the line segment connecting the two reference points S1 and S0 is set as a second virtual plane VP1. The point where a straight line passing through the center (0, 0, 0) of the image sensor 5 and the reference point S0 for tilt angle and curvature measurement intersects the second virtual plane VP1 is defined as a second reference point _PO '. . The distance L _xy in the xy plane is caused by the inclination angle Ψ of the measured surface 4a of the object 4 with respect to the second virtual plane VP1. The distance Lz in the direction of the optical axis _z is caused by convex or concave deformation of the measured surface 4a of the object 4 with respect to the second virtual plane VP1.

(Calculation of Ψ)
The inclination angle Ψ between the surface to be measured 4a (main surface of the semiconductor substrate) of the object 4 and the bottom reference (surface opposite to the main surface) is determined by a line segment connecting two points P _O ′ and S0 and two points P _O ' and the line segment connecting S1. Therefore, the tilt angle Ψ is obtained from the following equation using the distance R0 between the two points P _O ′ and S0 and the distance L _xy between the two points S1 and S0.
tan([psi])= _Lxy /(2R0).

(Calculation of Rs)
The radius of curvature of the spherical wave when the spherical wave light emitted from the reference point S0 for tilt angle and curvature measurement reaches the point P _O ' on the second virtual plane VP1 is the distance R0 between P _O ' and S0. . The radius of curvature of the spherical wave when the spherical wave light emitted from the reference point S1 for shape measurement reaches the point P _O ' on the second virtual plane VP1 is the distance R1 between P _O ' and S1. The optical path difference between the two spherical wave lights is generated by the deformation of the surface 4a (main surface of the semiconductor substrate) of the object 4 to be measured. Let the radius of curvature of the deformation be Rs. Also, R0 and R1 can be set to sufficiently large values compared to the size of the surface to be measured.

Here, if R1=R0±Δr, then Δr<<R0. The radius of curvature Rs of deformation can be expressed in a simple form that is intuitively understandable as in the following formula using the radiuses of curvature R0 and R1. The center position of Rs is on a line that passes through the midpoint of the line segment connecting S1 and S0 and is parallel to the line that passes through the center (0, 0, 0) of the image sensor 5 and S0.
1/Rs=1/R0−1/R1, Rs≈±R0 ² /Δr.

(Surface shape and three-dimensional shape of object 4)
The surface shape of the surface to be measured 4a with respect to the bottom surface of the object 4 is the planar distribution (RMS) of the phase of the complex amplitude inline hologram J ^V _OS , the radius of curvature (Rs) of the surface to be measured, and the inclination of the surface to be measured. It is obtained by synthesizing the angle (Ψ). Moreover, since the thickness D of the object 4 (thickness at the point P ₀ ′) can be obtained as described above, the three-dimensional shape of the object 4 can be measured. Focusing techniques can be set to locate any point in space. In the embodiment of the present invention, it is possible to detect errors on the _order of _nanometers . can be matched within the error obtained by adding the error due to the focus error detection and the error of the drive operation. The information on the tilt angle, the information on the curvature radius related to the convex or concave deformation of the substrate, and the information on the thickness and flatness of the silicon semiconductor substrate can be used for checking the quality of the substrate and determining the quality of the substrate. can.

It should be noted that the present invention is not limited to the above configuration, and various modifications are possible. For example, it is possible to adopt a configuration in which the configurations of the respective embodiments and modifications described above are combined with each other.

The novelty and superiority of the present invention over conventional techniques include the following: (1) High-speed measurement is possible by one-shot recording of light waves, (2) High-precision absolute flatness measurement of the surface 4a to be measured is possible. (3) Since no reference plane or collimator lens is used, a large aperture can be used for flatness measurement. (4) Flatness can be measured for the surface 4a to be measured having a wide range of reflection coefficients. (5) The surface to be measured. (6) The shape of the surface to be measured 4a of the object to be measured can be measured with reference to the bottom surface, and the shape of the surface to be measured 4a can be measured with respect to the bottom surface of the object to be measured. The curvature, parallelism and thickness of the surface 4a can be measured.

Due to the advantages described above, the present invention can be applied to a wide range of applications in the fields of optics, digital holography, optical measurement, interferometry, and fine shape measurement, making use of these advantages. From the viewpoint of technical application, it can be used in fields such as precision measurement, nanotechnology, substrate shape measurement, semiconductor substrate inspection, and optical component inspection. Specific usage examples include surface shape measurement of thin glass substrates, photomasks, large wafers, etc., surface shape measurement of optical components, measurement of industrial reference planes, and the like.

1 surface profile measuring device 10 data acquisition unit 12 image reconstruction unit 13 complex amplitude hologram generation unit 14 calculation reference beam hologram generation unit 15 object beam hologram generation unit 16 reconstruction object beam hologram generation unit 17 reference point detection unit 18 analysis beam hologram generation unit 19 shape measuring unit 2 optical system 27 condenser lens 27a pupil plate 27b imaging lens 28 concave mirror 28a pupil plate 28b imaging lens 3 beam combiner (cube beam splitter)
4 object 4a surface to be measured 40 plane-parallel substrate 40a reference plane of plane-parallel substrate 5 image sensor 50 light receiving surface 6 data storage unit 7 sample _stage C correlation function _ILR , ^Ij _LR reference beam off-axis hologram ^IOR , _IjOR Object beam off-axis hologram _JOL complex amplitude inline hologram of object beam _JVOS , ^JjVOS complex amplitude inline _hologram for measurement ^L inline spherical wave reference beam L _xy reference point S1 and reference point S0 in the _xy plane Distance L Distance between _z reference point S1 and reference point S0 in the direction of optical axis z _O Object light Pa Irradiation point on virtual plane P _L Condensing point of inline spherical wave reference light PO reference point _PO ' Second reference point P _R Condensing point of off-axis reference light Q Illumination light R Off-axis reference light R0 Curvature radius at reference point Po of spherical wave emitted from reference point S0 R1 Curvature radius at reference point Po of spherical wave emitted from reference point S1 Rs Curvature radius of the surface 4a to be measured RMS Root mean square
S0 reference point for tilt angle and curvature measurement S1 reference point for shape measurement (reference beam point light source)
VP virtual plane VP1 second virtual plane fp virtual point light source (probe function)
g object beam hologram h0 evaluation hologram h ^V reconstructed object beam hologram j _L inline reference beam hologram s ^V spherical wave beam hologram α _O tilt angle ρ distance from image sensor to focal point of inline spherical wave reference beam λ _B modulation wavelength λ _j , λ ₁ , λ ₂ wavelength θ _B modulation phase Ψ Inclination angle of measured surface 4a with respect to reference plane

Claims (12)

In a surface shape measuring device using holography,
Two lights: an object light (O) that is reflected light of the spherical wave illumination light (Q) that illuminates the surface to be measured of the object, and an inline spherical wave reference light (L) that is inline with the object light (O). as an object beam off-axis hologram (I _OR ) and a reference beam off-axis hologram (I _LR ) using the off-axis reference beam (R), respectively;
an image reproduction unit that performs calculation processing for reproducing the image of the surface to be measured using the data acquired by the data acquisition unit and acquiring surface shape data of the surface to be measured;
The data acquisition unit
an image sensor that converts light intensity into an electrical signal and outputs it as hologram data;
a sample table that holds the object with the surface to be measured facing the light receiving surface of the image sensor;
An illumination light condensing point (P _Q ), which is the condensing point of the spherical wave illumination light (Q), and a reference light condensing point (P _L ), which is the condensing point of the inline spherical wave reference light (L) , are mirror images of each other with respect to a virtually set virtual plane (VP), and the inline spherical wave reference light (L) passes through the virtual plane (VP) and is incident on the image sensor. an optical system for obtaining a hologram;
a position detection unit that detects the position of a point on the surface to be measured in an area including the front and back of the virtual plane (VP) in the optical axis direction of the image sensor and outputs the information as reflection position information;
a position adjustment device that moves the sample stage so that the detected point is included in the virtual plane (VP) based on the reflection position information;
The image reproducing unit
Using the fact that the light emitted from the reference light condensing point (P _L ) is a spherical wave, the light wave of the object light (O) is obtained from the data of the two types of off-axis holograms (I _OR , I _LR ). an object beam hologram generator for generating an object beam hologram (g) representing
a reconstructed object light hologram generation unit for optically propagating and rotationally transforming the object light hologram (g) to generate a reconstructed object light hologram (h ^V ) on the virtual plane (VP);
a reference point detection unit that performs optical propagation conversion on the object light hologram (g) to detect a position where the object light (O) is focused and sets the position as a reference point (S1) for shape measurement;
an analysis light hologram generator that analytically generates a spherical wave light hologram (s ^V ), which is a hologram in the virtual plane (VP) of spherical wave light emitted from the reference point (S1);
a shape measuring unit that obtains a height distribution of a surface to be measured of the object from a surface distribution of the phase difference between the reproduced object optical hologram (h ^V ) and the spherical wave optical hologram (s ^V ). Surface shape measuring device. The data acquisition unit is arranged immediately before the image sensor, combines the object light (O) or the in-line spherical wave reference light (L) and the off-axis reference light (R), and outputs the data to the image sensor. Equipped with a beam combiner consisting of a cube-shaped beam splitter for incidence,
The image reproducing unit calculates the light propagation of the light passing through the beam _combiner by a plane wave expansion method considering the refractive index of the beam combiner, thereby obtaining a an in-line reference light hologram (j _L ) representing a light wave that passes through the beam combiner and reaches the light-receiving surface of the image sensor, the light wave corresponding to the in-line spherical wave reference light (L) at the light-receiving surface. 2. The surface profile measuring device according to claim 1, wherein the surface profile measuring device generates a The position detection unit is
a detection light source that emits a light beam for position detection that is reflected from the surface to be measured;
a photodetector that detects the reflected light of the light beam and outputs an electrical signal in response to the detection;
a detection optical system that irradiates a predetermined irradiation point on the virtual plane (VP) with a light beam from the detection light source and causes the reflected light of the light beam irradiated to the irradiation point to enter the photodetector;
Based on the electrical signal output from the photodetector, a shift signal indicating a distance between the position where the light beam is reflected and the virtual plane (VP) is generated, and the position adjustment is performed as the reflected position information. 3. The surface profile measuring device according to claim 1, further comprising a signal output unit for outputting to the device. The photodetector has a plurality of detection elements,
4. The surface shape measuring apparatus according to claim 3, wherein the signal output unit uses a difference signal of signals output from each of the plurality of detection elements as the deviation signal. The position detection unit is
a detection light source that emits a modulated light beam used for detecting the position of the surface to be measured;
two photodetectors for detecting the light beam;
a beam splitter for separating the light beam from the detection light source into two light beams;
One of the separated light beams is incident on one of the photodetectors, the other of the separated light beams is propagated toward the surface to be measured, and the reflected light from the surface to be measured is detected as the light beam. a detection optical system that is incident on the other side of the photodetector;
generating a shift signal indicating the distance between the position of the surface to be measured and the position of the virtual plane (VP) based on the time difference between the detection of the separated light beams by the two photodetectors, respectively; 3. The surface shape measuring apparatus according to claim 1, further comprising a signal output unit for outputting the reflected position information to the position adjustment device. The optical system for obtaining the hologram includes a condenser lens for condensing the object light (O) and the inline spherical wave reference light (L), and is arranged at a condensing position by the condenser lens to reduce the amount of passing light. a limiting pupil plate and an imaging lens arranged in combination with the pupil plate to image the object light (O) and the in-line spherical wave reference light (L) on the image sensor; 6. The surface shape measuring apparatus according to any one of claims 1 to 5, characterized in that: The optical system for obtaining the hologram includes a concave mirror that converges the object light (O) and the inline spherical wave reference light (L), and a pupil plate that is arranged at a converging position of the concave mirror to limit the amount of light passing through. and an imaging lens arranged in combination with the pupil plate to form an image of the object light (O) and the inline spherical wave reference light (L) on the image sensor. The surface shape measuring device according to any one of claims 1 to 5. In a surface shape measurement method for measuring the shape of a surface to be measured of an object using holography,
A reference light condensing point (P _L ), which is a condensing point of the inline spherical wave reference light (L), is arranged on the optical axis of the image sensor, and the spherical wave illumination light (Q) is located at a position off the optical axis. An illumination light condensing point (P _Q ) is arranged, and a line segment connecting the reference light condensing point (P _L ) and the illumination light condensing point (P _Q ) is vertically bisected. Set a virtual plane (VP), which is the plane to
setting a reference point (P _O ) at the intersection position of the virtual plane (VP) and the optical axis;
holding a plane-parallel substrate having a reference plane on a sample stage so that the reference plane faces the light receiving surface of the image sensor;
Adjusting the sample table so that the reference plane is in contact with the virtual plane (VP);
holding the object instead of the parallel plane substrate on the sample table so that the surface to be measured faces the light receiving surface of the image sensor;
adjusting the position of the sample table so that the surface to be measured is in contact with the virtual plane (VP);
obtaining data of object light (O), which is reflected light of the spherical wave illumination light (Q) from the surface to be measured, as an object light off-axis hologram (I _OR ) using the image sensor;
Data of the in-line spherical wave reference light (L) passing through the virtual plane (VP) and incident on the image sensor in a state in which the parallel plane substrate and the object are not arranged are captured using the image sensor. acquired as a reference beam off-axis hologram (I _LR ),
Object light representing a light wave of the object light (O) from the data of the two types of off-axis holograms (I _OR , I _LR ) by calculation processing using the inline spherical wave reference light (L) being spherical wave light generating a hologram (g);
optically propagating and rotationally transforming said object beam hologram (g) by computational processing to produce a reconstructed object beam hologram (h ^v ) in said virtual plane (VP);
Light propagation conversion is performed on the object light hologram (g) by calculation processing to detect the position where the object light (O) is condensed, and the position is converted to the illumination light condensing point with respect to the virtual plane (VP). Considered as a mirror image point of (P _Q ) and set as a reference point (S1) for shape measurement,
analytically generating a spherical wave light hologram (s ^v ) which is a hologram in said virtual plane (VP) of spherical wave light emitted from said reference point (S1);
A surface profile measuring method, wherein a height distribution of a surface to be measured of the object is obtained from a surface distribution of a phase difference between the reproduced object optical hologram ( ^hv ) and the spherical wave optical hologram ( ^sv ). The adjustment of the sample table so that the reference plane is in contact with the virtual plane (VP) includes:
The data of the in-line spherical wave reference beam (L) passing through the virtual plane (VP) and incident on the image sensor in the state where the parallel plane substrate is not arranged is converted into a reference beam off-axis hologram (I _LR ). and get as
holding the plane-parallel substrate on the sample stage and acquiring data of reflected light of the spherical wave illumination light (Q) from the reference plane as an object light off-axis hologram (I _OR );
The method is characterized in that the position and inclination of the sample stage are changed so as to reduce a change in surface distribution of the phase difference between the object light off-axis hologram (I _OR ) and the reference light off-axis hologram (I _LR ). The surface shape measuring method according to claim 8. The adjustment of the sample table so that the surface to be measured is in contact with the virtual plane (VP) includes:
A position detection light beam is irradiated toward one point on the virtual plane (VP), and the reflected light is received, whereby the distance between the position where the light beam is reflected and the virtual plane (VP) 10. The method according to claim 8 or 9, characterized in that active autofocus technology is used to generate a deviation signal according to the deviation signal and to move the surface to be measured so that the deviation signal becomes small. surface profile measurement method. The setting of the reference point (S1) for shape measurement is
A probe function ( _fp ) is used to detect the condensing point in the plane of the evaluation hologram (h0), and the evaluation hologram (h0) is experimentally propagated in the optical axis direction by light propagation calculation, and the probe function ( 11. The method according to any one of claims 8 to 10, wherein a condensing point in the optical axis direction is detected by calculating a correlation function with fp), and the condensing point is set as the reference point (S1). or the surface shape measurement method according to the item 1. Data of the object light (O) and the in-line spherical wave reference light (L) are converted to the data of the object light (O) and the in-line spherical wave reference light (L) for each wavelength (λ ₁ , λ ₂ ) by light of different wavelengths (λ _j , j=1, 2). obtained as off-axis holograms of the kind (I ^j _OR , I ^j _LR , j=1, 2),
Configured as a ratio of the reconstructed object light hologram (h _j ^V , j=1, 2h ^V ) and the spherical wave light hologram (s _j ^V , j=1, 2) for each wavelength (λ ₁ , λ ₂ ) to generate a measurement hologram (J _j ^V _OS =h _j ^V /s _j ^V , j=1,2),
generating a modulated wave (HW=J ₁ ^V _OS /J ₂ ^V _OS ) which is a result of heterodyne transformation for obtaining a ratio of the two measurement holograms (J _j ^V _OS , j=1, 2); Using the modulation wavelength (λ _B =λ ₁ λ ₂ /(λ ₂ −λ ₁ )) and modulation phase distribution (θ _B (x′, y′)=θ ₁ −θ ₂ ) included in (HW), 12. The surface shape measuring method according to claim 8, wherein a height distribution on the surface to be measured of said object is obtained.

Images