JP2013213695A - Shape measurement device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a three-dimensional shape measurement device having confocal resolution in a lateral direction and resolution of a wavelength order or less in a vertical direction.SOLUTION: A shape measurement device includes: a confocal optical system 1A including a coherent light source 12, a scanning optical system 21 and a first detection optical system 31; and a movable mechanism 29 changing a height up to a surface of a sample S. The confocal optical system 1A includes a beam splitter 24 branching a luminous flux and a reference mirror 26 reflecting one luminous flux branched by the beam splitter 24 and making the luminous flux incident on the first detection optical system 31. The shape measurement device detects interference outputs of measuring heights in at least three portions adjacent to a measuring object portion on the sample S, repeatedly performs an operation for determining a peak position and amplitude of an interference output in the vicinities of the measuring heights in the at least three portions for each measuring height, and calculates height and quantity of light emitted at each measuring object portion on the sample S on the basis of a maximum value of the amplitude of the obtained inference pattern and the height.

Description

  The present invention relates to a shape measuring apparatus, and more particularly to a shape measuring apparatus that measures or observes a fine three-dimensional shape using a scanning confocal optical system.

As a device for measuring / observing a fine three-dimensional shape, a device using a white light interference optical system (Patent Documents 1 and 2) and a device using a confocal optical system (Non-Patent Document 1) are known. .
Both devices measure the height direction by irradiating the sample with illumination light while shifting the distance (measurement height) between the optical system and the sample (object) at a predetermined interval. This is done by measuring the amount of light.

  In an apparatus using a white light interference optical system, a sample and a reference surface are irradiated with a low-coherence light beam having a wide spectral component in the visible light range, and the light beams reflected from the sample and the reference surface are combined and incident on an image sensor. . On each pixel of this image sensor, the light beam reflected from the sample and the light beam reflected from the reference surface overlap each other. However, since the coherence of the light source is low, an interference waveform appears only in a region where the respective optical path lengths are approximately the same. The height of the part of the sample corresponding to each pixel is measured from the position of the envelope peak of the interference waveform, and an intensity image of the part of the sample is obtained from the interference amplitude at the peak.

The distance between the sample and the optical system is sampled at a predetermined interval. Here, in order to obtain the peak position, several methods have been proposed based on the vertical scanning white light interference (VSI: Vertical Scanning Interferometry) method (see Patent Document 3, etc.).
On the other hand, there are several methods for an apparatus using a confocal optical system, but the confocal optical system is configured using a laser as a light source, and the laser spot position formed on the sample is scanned within the measurement plane. A method of obtaining a two-dimensional image by detecting the amount of reflected light at each sampling point is widely used. When this confocal optical system is used, the resolution (lateral resolution) and contrast in the measurement surface are high compared to the normal white light method, and particularly when an objective lens with a large NA is used, A resolution in the height direction (vertical resolution) at the wavelength level can be obtained.

Special table 2009-509150 Japanese National Patent Publication No. 09-503065 Japanese Patent No. 3511097

Satoshi Kawada, Shigeo Minami, "Optical Scanning Microscope", Optics, 1989, Vol. 18, No. 8, p. 380

By the way, the apparatus using the conventional white light interference optical system has a problem that the lateral resolution is low as compared with the apparatus using the confocal optical system.
On the other hand, the apparatus using the confocal optical system has a problem that when the NA of the objective lens is small, the vertical resolution is remarkably reduced in proportion to the square of NA.
As described above, the features differ depending on each method, and there is a problem that a shape measuring apparatus having both high lateral resolution and high vertical resolution cannot be obtained.

  The present invention has been made in view of the above circumstances, and in the case of an apparatus using a confocal optical system, it is possible to increase both the horizontal resolution and the vertical resolution. It is an object of the present invention to provide a three-dimensional shape measuring apparatus having a resolution of 2 nm and a resolution of the order of the wavelength or less in the longitudinal direction regardless of the focal length of the objective lens.

  As a result of intensive studies to solve the above-mentioned problems, the present inventors, as a result of the first measurement for branching the light beam emitted from the first light source to the scanning optical system, for the shape measuring apparatus using the confocal optical system. In addition to providing a branching means, a reflecting means for reflecting one of the light beams branched by the first branching means and entering the detection optical system is provided, and interference of three or more measurement heights adjacent to the measurement target part on the sample is provided. Obtaining the peak position (height) and amplitude of the interference output (output light quantity) in the vicinity of these three or more measurement heights by repeatedly detecting the output (output light quantity) for each measurement height. It is found that both the horizontal resolution and the vertical resolution can be improved by calculating the height and the amount of emitted light of each measurement target portion of the sample based on the maximum amplitude of the interference pattern and its height. To complete the present invention .

That is, the shape measuring apparatus of the present invention includes a first light source, a scanning optical system that converges on the surface of the sample while scanning the light beam emitted from the first light source, and the sample. A shape measuring apparatus comprising a confocal optical system including a detection optical system for detecting an actuated light beam, and variable means for changing the height of the sample to the surface,
The confocal optical system is provided with a first branching unit for branching the light beam, and a reflecting unit for reflecting one light beam branched by the first branching unit and entering the detection optical system,
The operation of detecting the interference output at three or more measurement heights adjacent to the measurement target site on the sample and obtaining the peak position and amplitude of the interference output near the three or more measurement heights for each measurement height. The measurement is repeated every time, and the height and the amount of emitted light of each measurement target portion of the sample are calculated based on the maximum amplitude value and the height of the obtained interference pattern.

The range for performing the phase analysis of the height is set to exceed one period of the interference pattern, and the relative phase from the start point of each measurement point that exceeds one period of the interference pattern from the start point of the range, It is preferable to perform the phase analysis corresponding to the range of one cycle.
The peak position and amplitude of the interference pattern are preferably calculated using a phase shift (PSI) method.

A second light source having lower coherence than the light source, and a second optical system for guiding a second light beam generated from the second light source onto the surface of the sample via the objective lens of the scanning optical system; And an imaging lens that forms an image of the second light flux that interacts with the sample, and an imaging device that is disposed on the imaging surface of the imaging lens, and the reflection means is shared with the confocal optical system. A coherence optical system, and an optical distance from the reflecting means to the branching means is substantially equal to an optical distance from the branching means to the focal point of the objective lens, and corresponds to a position for measuring the height of the sample. From the height information at the position of the low-coherence image, the height at which the intensity of the pixel with respect to the measurement site of the low-coherence image reaches a peak is obtained. From the interference output data it is preferable to determine the amount of light emitted from the peak position in the confocal optical system.
The first branching unit preferably has a branching ratio to the reflecting unit of 1/3 or less.

At the time of calibration or measurement start, four or more positions are selected from the height displacement for one period of the interference pattern obtained from the nominal wavelength of the first light source, and interference outputs are obtained, and these interferences are obtained. It is preferable to obtain the actual wavelength from the output data.
The interference output data acquired at the four or more positions is preferably only one point or one line data in the scanning plane on the surface of the sample.

  According to the shape measuring apparatus of the present invention, the confocal optical system is provided with the first branching unit for branching the light beam, and the reflection for reflecting one of the light beams branched by the first branching unit and entering the detection optical system. Means for detecting interference outputs at three or more measurement heights adjacent to the measurement target site on the sample, and performing operations for obtaining peak positions and amplitudes of interference outputs near the three or more measurement heights. The measurement is repeated at each measurement height, and the height and the amount of emitted light of each measurement target part of the sample are calculated based on the maximum amplitude value and the height of the obtained interference pattern. Can be increased. Accordingly, it is possible to realize a three-dimensional shape measuring apparatus having a confocal resolution in the horizontal direction and having a resolution of the wavelength order or less regardless of the focal length of the objective lens in the vertical direction.

It is a schematic block diagram which shows the shape measuring apparatus of one Embodiment of this invention. It is a figure which shows various interference waveforms in the shape measuring apparatus of one Embodiment of this invention. It is a flowchart which shows the procedure which acquires data with the shape measuring apparatus of one Embodiment of this invention. It is a flowchart which shows the procedure which calculates | requires the height of an object part, and emitted light quantity from the data acquired by the shape measuring apparatus of one Embodiment of this invention.

The form for implementing the shape measuring apparatus of this invention is demonstrated.
This embodiment is specifically described for better understanding of the gist of the invention, and does not limit the present invention unless otherwise specified.

FIG. 1 is a schematic configuration diagram showing a shape measuring apparatus according to an embodiment of the present invention. The shape measuring apparatus 1 is a confocal optical system 1A, a low coherence optical system 1B, and an operation control of the shape measuring apparatus 1. And a control device 2 that executes signal processing, and a display (not shown) for displaying a fine three-dimensional image from the shape measuring device 1, information on measurement, and the like.
The confocal optical system 1A converges the first illumination optical system 11 that emits a coherent light beam and the light beam emitted from the first illumination optical system 11 on the surface of the sample S while being scanned. A scanning optical system 21 and a first detection optical system 31 that detects light generated from the sample S through a pinhole (confocal stop) 34 are provided.

The first illumination optical system 11 includes a coherent light source (first light source) 12 that emits a coherent light beam, and a collimator lens 13 that collimates the light beam emitted from the coherent light source 12.
The coherent light source 12 is a laser light source that emits a coherent light beam, for example, a semiconductor laser (LD) that emits violet laser light. The coherent light source 12 is driven by a laser drive circuit (not shown) controlled by the control device 2.

The scanning optical system 21 includes a first scanning mirror 22, an optical system 23 that forms an optical path by combining a pupil relay lens (fθ lens), a scanning mirror, and the like, a beam splitter (first branching unit) 24, and an objective lens. 25, a reference mirror (reflecting means) 26, and an objective lens 27.
The first scanning mirror 22 is constituted by a resonance mirror. By rotating the first scanning mirror 22 around the Y axis perpendicular to the paper surface, the reflected light beam is in the X-axis direction (one direction horizontal to the paper surface). To deflect. The first scanning mirror 22 can be rotated by a mirror driving unit (not shown), and the rotation driving is controlled by the control device 2.

The optical system 23 is an optical system that combines a pupil relay lens (fθ lens), a scanning mirror (not shown), and the like. In this optical system 23, it is possible to deflect the light beam in the Y-axis direction (one direction perpendicular to the paper surface) by rotating a scanning mirror (not shown) around the X-axis horizontal to the paper surface.
The reference mirror 26 is arranged at a position where the optical distance from the reference mirror 26 to the beam splitter 24 is substantially equal to the optical distance from the beam splitter 24 to the focal plane of the objective lens 27, and is branched by the beam splitter 24. The light beam collected by 25 is reflected in the original direction as a reference light beam and is incident on the first detection optical system 31.

  An interference objective lens in which the functions of the beam splitter 24, the objective lens 25, and the reference mirror 26 relating to the generation of the reference light beam are incorporated in the objective lens 27 is commercially available. The interference objective lens includes a Mirau type and a Michelson type, and both can be used. The Michelson type is suitable for an objective lens having a long working distance (WD) such as a low magnification. Therefore, when these interference objective lenses are used, it is not necessary to arrange the beam splitter 24, the objective lens 25, and the reference mirror 26 at the illustrated positions unless the low coherence optical system 1B is used.

  The stage 28 mounts the sample S, and is fixed on a movable mechanism (variable means) 29 so that the height of the surface of the sample S can be changed by moving the sample S along the optical axis. By changing the height of the surface of the sample S by the movable mechanism 29, the vertical scanning of the light beam can be performed. Instead of moving the movable mechanism 29, the objective lens 27 may be fixed to the movable mechanism 29 and the distance from the sample S may be variable.

In this way, the first scanning mirror 22 performs an operation of deflecting the light beam in the X-axis direction on the surface of the sample S, thereby performing one-dimensional scanning (scanning in the X-axis direction) on the surface (XY plane) of the sample S. Is possible.
Further, when the operation of deflecting the light beam by the first scanning mirror 22 in the X-axis direction and the operation of deflecting the light beam by the optical system 23 in the Y-axis direction are combined, two-dimensional scanning of the surface (XY plane) of the sample S is performed. Is possible.
The light beam deflected in the X-axis direction by the first scanning mirror 22 passes through the beam splitter 24, enters the objective lens 27, and is condensed on the sample S.

The first detection optical system 31 includes a beam splitter 32, a pinhole lens 33, a pinhole (confocal stop) 34, and a photodetector 35.
The beam splitter 32 is disposed between the first illumination optical system 11 and the scanning optical system 21 and separates the light beam applied to the sample S and the light beam generated from the sample S.

The pinhole lens 33 condenses the light incident from the beam splitter 32 in the pinhole 34.
The photodetector 35 is an element that detects light incident through the pinhole 34, converts the amount of received light into an electrical signal, and outputs the electrical signal. As the photodetector 35, a photodiode or an avalanche photodiode is used. And photomultiplier tubes.
The photodetector 35 outputs the electrical signal to the control device 2, and in the control device 2, the calculation unit 62 creates an electrical signal of a display image based on the electrical signal from the photodetector 35, and displays a display (not shown). ) To display the display image.

In this confocal optical system 1 </ b> A, the light beam emitted from the coherent light source 12 is converted into parallel light by the collimator lens 13, passes through the beam splitter 32, and enters the first scanning mirror 22 of the scanning optical system 21. .
The light beam incident on the scanning optical system 21 is deflected in the X-axis direction in the XY plane of the surface of the sample S by the first scanning mirror 22, passes through the beam splitter 24, and arbitrarily passes on the surface of the sample S by the objective lens 27. The position is focused and irradiated.
For example, two-dimensional scanning of the surface of the sample S (XY plane) is possible by repeatedly performing the operation of deflecting the light beam in the X-axis direction on the surface of the sample S along the Y-axis direction.

  In this confocal optical system 1A, interference outputs (output light amounts) at three or more measurement heights adjacent to a measurement target site on the sample S are detected, and interference outputs (output light amounts) near these three or more measurement heights ( The operation for obtaining the peak position (height) and amplitude of the output light amount) is repeated for each measurement height, and each measurement target part of the sample S is based on the maximum value and the height of the obtained interference pattern amplitude. And the amount of emitted light can be calculated.

On the other hand, the low-coherence optical system 1B is configured to receive a second illumination optical system (second optical system) 41 that irradiates the sample S with low-coherence (white) light and reflected light from the sample S irradiated with this light. And a second detection optical system 51 for detection.
In the low coherence optical system 1B, the beam splitter 24, the objective lens 27, the objective lens 25, and the reference mirror 26 are shared with the confocal optical system 1A, and the optical path is branched by the beam splitter 24 described above.

  The second illumination optical system 41 includes a white light source (second light source) 42 that emits white illumination light, a collector lens 43, and a half mirror 44. The optical path is branched by the half mirror 44, and the light flux returned from the sample S and the reference mirror 26 is reflected by the half mirror 44 and reaches the second detection optical system 51.

Examples of the white light source 42 include a halogen lamp, a xenon lamp, a metal halide lamp, and an LED.
The collector lens 43 collects white light emitted from the white light source 42 to form an image. At that time, in the white light interference optical system, the image of the light source is often formed on the sample S surface and the reference mirror 26 through the objective lens, but it is not always necessary to do so. For example, it is possible to set so that the image of the light source is formed near the pupil planes of the objective lens 27 and the objective lens 25.
The half mirror 44 is an optical element that transmits light emitted from the white light source 42 and reflects light generated from the sample S.

The second detection optical system 51 includes a half mirror 44, an imaging lens 52, and an imaging element 53 disposed on the imaging surface of the imaging lens 52.
The half mirror 44 divides the second illumination optical system 41 and the second detection optical system 51, and is arranged on the optical axis of the second illumination optical system 41 so that the light reflected by the sample S can be reflected. Reflected toward the imaging lens 52 side. The imaging lens 52 condenses incident light on the imaging surface of the imaging element 53 and forms an image of the sample S on the imaging element 53. The image sensor 53 forms an image of the sample S by detecting incident light.

  In the low coherence optical system 1B, the white illumination light emitted from the white light source 42 passes through the collector lens 43 and the half mirror 44 and enters the beam splitter 24. Thereafter, the white illumination light is reflected by the beam splitter 24 and condensed on the sample S by the objective lens 27.

  The light reflected by the surface of the sample S passes through the objective lens 27, is reflected by the beam splitter 24, is further reflected by the half mirror 44, is collected by the imaging lens 52, and is collected by the imaging element 53. An image is formed on the imaging surface. Thereby, an observation image (non-confocal image) of the sample S is formed. At the same time, the illumination light transmitted through the beam splitter 24 is reflected by the reference mirror 26, is again transmitted through the beam splitter 24, is reflected by the half mirror 44, and enters the image sensor 53 through the imaging lens 52.

  By using this low-coherence optical system 1B, the intensity of this low-coherence image peaks from the output of each measurement height at the position (pixel) of the low-coherence image corresponding to the position of the sample S that is the height measurement target. And the peak value of the output light quantity, that is, the emitted light quantity of the target portion of the sample S can be obtained from the output data of the confocal optical system 1A for the height around the peak height. .

  The control device 2 includes a control unit (not shown) that controls the two-dimensional scanning of the confocal optical system 1A, a control unit 61 that controls the height of the stage 28, the confocal optical system 1A, and the low coherence optical system 1B. And an image receiving circuit 63 that inputs an output signal from the image sensor 53 and sends the output signal to the computing unit 62.

Next, the operation of the shape measuring apparatus 1 will be described.
First, the range of vertical scanning of the stage 28 is determined, and the stage 28 is set at the start position of this vertical scanning. Here, data acquisition is performed according to the type of coherent interference waveform and whether to use white light interference together.

(1) When Corresponding to Coherent Interference Waveform (1) in FIG. 2 First, data in a predetermined vertical scanning range is acquired while moving the position of the stage 28 in the vertical direction at regular intervals, and then acquired. Using the data, the height of each target part and the amount of emitted light are calculated (estimated) by the PSI method or the like.
In the PSI method, the interference output data is basically obtained by dividing the interference period into fixed intervals. For example, when using data of four or more locations, a certain amount of error is allowed in the division position. ing. The details are described in Reference 1, “Kieran G. Larkin, JOSA, Vol. 13, No. 4, p. 832, 1996”.
In addition, when the optical system has a shallow depth of focus, a change in received light intensity due to defocusing occurs within a distance equivalent to the wavelength, so it is necessary to perform processing within an interval of about one cycle of the interference waveform. is there.
For example, when the 5-step method using data acquired from five locations is used, the data acquisition interval is preferably about 1/8 of the laser wavelength of the light source used. As a result, four measurement positions are included in one cycle of the interference output change.

Therefore, when the data acquisition in the predetermined vertical scanning range is completed, the bias and phase of the interference waveform are obtained by the following analysis.
Here, for the photodetector 35, the following analysis is performed only in the range of the measurement height where the detected light amount is a certain ratio (for example, 50%) or more with respect to the peak.
Note that when performing the above-described determination at a certain ratio, the detected light amount fluctuates due to interference. Therefore, it is preferable to smooth the measurement data by taking a moving average for each measurement.
When the data acquisition in the predetermined vertical scanning range is completed, the bias and phase of the interference waveform are obtained.

ただし、式（１）中、ｚは原点からの縦（光軸）方向の測定位置、φは原点からのピーク位置（位相）、λは波長、Ａは振幅、Ｂはバイアス、ｎは屈折率（既知）である。 The interference waveform S (z) can be expressed by the following equation (1).

In the formula (1), z is a measurement position in the longitudinal (optical axis) direction from the origin, φ is a peak position (phase) from the origin, λ is a wavelength, A is an amplitude, B is a bias, and n is a refractive index. (Known).

In the shape measuring apparatus 1, since the light beam is reflected by the sample S, the optical path length change is twice as large as the height displacement (z).
The four unknowns A, B, λ, and φ can be obtained by obtaining the interference output S (z) for each of the height displacements (z) at four or more measurement heights.
If the above calculation is repeated by shifting the data position (measurement height) to be used by a certain number, the position of the interference peak near each position (measurement height) (from the peak position of the envelope curve to the interference pattern period) And the interference output S (z) is obtained. Therefore, if the calculation is repeated by shifting one by one, the amount of calculation increases, but it is possible to cope with a sudden change in amplitude.

Subsequently, the light quantity of the emitted light from the sample S is calculated | required as follows about the position (peak position of an envelope) where this interference output S (z) is the largest.
Here, assuming that the amplitude of the reference light is A _r and the amplitude of the light from the sample S is A ₀ , the bias B and the amplitude A in equation (1) are expressed by the following equations (2) and (3). Can do.
B = A _r ² + A ₀ ² (2)
A = 2A _r · A ₀ (3)
However, since the amount of reference light (A _r ² ) is constant and can be measured in advance, the amount of reference light from the sample is subtracted from the bias using equation (2). The amount of light (A ₀ ² ) can be obtained. Further, since the amplitude A _r of the reference light is constant, the amplitude A ₀ of the light from the sample S can be regarded as proportional to A, therefore, by obtaining the A of the formula (3), as a relative value A ₀ can be obtained from the amplitude of light from the sample S.

In a normal reflection type confocal optical system that does not use interference, if the detected light quantity characteristic with respect to defocus (corresponding to z above) is taken, the range of z that is half the light quantity with respect to the peak light quantity is When the NA of the lens is as high as about 0.9 or higher, the order of wavelengths is generally reached. Therefore, the resolution in the z direction of the confocal optical system is 0.1 μm to 1 μm when visible light is used.
In contrast, when the peak position is obtained by analyzing the phase as described above, it is possible to obtain a resolution of 1/100 to 1/10000 of the interference period, depending on the SN ratio of the confocal optical system. it can. This resolution corresponds to about 0.1 nm to 0.01 μm in visible light.

  By introducing interference into the confocal optical system, the change in the amount of light becomes steeper compared to a normal optical system. Therefore, it becomes easier to specify the interference cycle showing the peak, and the sample S having a large height change is used. In contrast, the height can be detected with high resolution.

(2) Corresponding to the coherent interference waveform (2) in FIG. 2 When the NA of the objective lens is relatively small and the depth of focus is deep, each pixel of the photodetector 35 is constant with respect to the peak light quantity. The range of the above ratio becomes wider. Along with this, the number of interference periods included in the period also increases, and the light amount change for each interference period becomes small particularly near the peak.
Here, when the depth of focus is significantly deeper than the wavelength, for example, when the wavelength is 1 μm or less and the depth of focus is 50 μm or more, the amplitude and the bias are sufficiently smaller than the depth of focus and 1 / of the wavelength so that the amplitude and the bias can be regarded as constant. A predetermined number of data is acquired in an interval longer than 2, for example, 2 μm, and the same processing as in the above (1) is performed.

ただし、式（４）中、ｚは原点からの縦（光軸）方向の測定位置、φは原点からのピーク位置（位相）、λは波長、Ａは振幅、Ｂはバイアス、ｎは屈折率（既知）、Ｎは整数である。 However, when the range of three or more data to be processed exceeds one cycle, the following equation (4) is obtained.

In equation (4), z is the measurement position in the longitudinal (optical axis) direction from the origin, φ is the peak position (phase) from the origin, λ is the wavelength, A is the amplitude, B is the bias, and n is the refractive index. (Known), N is an integer.

In this case, since the phase cannot be uniquely determined, it is necessary to adopt N closest to the average position of the measured data range. Also in this formula (4), B and A can be regarded as constant, so there is no problem at all.
Next, the calculation is performed by shifting the data range one by one, and the phase and amplitude (A) in the vicinity of the position where the value of A is the largest are adopted.
When using the three-step method, if the change in the output light quantity is large at the start of measurement, take data at four or more locations within the interference period, obtain the wavelength accurately based on this data, and then use the interference It is preferable to acquire data at each step of 0 degree + 360 degrees × n, 90 degrees + 360 degrees × n, 180 degrees + 360 degrees × n (where n is an integer) with respect to the cycle.
This method can increase the speed even when the longitudinal scanning range becomes long.

(3) When the depth of focus is very deep in the coherent interference waveform (3) of FIG. 2 When the depth of focus is very deep and the coherent interference waveform (3) is obtained, the amplitude and bias are almost constant, so the envelope It becomes difficult to accurately determine the line peak.
Therefore, the envelope peak position is obtained by the low coherence interference method (VSI method) using the low coherence optical system 1B together.
When the low coherence optical system 1B is used, it is necessary to accurately match the optical distance from the beam splitter 24 to the surface of the sample S and the optical distance from the beam splitter 24 to the reference mirror 26. The reason is that an interference pattern cannot be obtained for light having a short coherence length unless the optical path length is equal to or less than the coherence length (about several μm).

In this case, when acquiring the interference output data (confocal image) of the confocal optical system 1A and the interference output partial data of one point or one line of the confocal optical system 1A, the low coherence image is acquired at the same stage position. Do.
When the height is obtained from the data of the low coherence optical system 1B using the VSI method, it is necessary to set the measurement height step more finely in order to obtain the height with high accuracy. In that case, the low-coherence image may be acquired not only when the confocal image is acquired but also at a height in between.
When acquiring a low coherence image, the output of the coherent light source 12 such as a semiconductor laser of the confocal optical system 1A is turned off, and when acquiring interference output data of the confocal optical system 1A, a shutter or the like is used to start from the low coherence optical system 1B. It is necessary to prevent the output light beam from entering the confocal optical system 1A.

The data acquisition procedure in this case will be described with reference to FIG.
First, the stage 28 is set at the vertical scanning start position (step S1).
Next, after starting vertical scanning, confocal data of one point or one line is acquired at four heights for pixels or lines in which a certain amount of light is first detected, and a wavelength is calculated from these confocal data. (Step S2).

  If the wavelength can be calculated in this way, a low coherence image is acquired (step S3), and then it is determined whether or not a confocal image is acquired from the set confocal image acquisition interval. (Step S4) When a confocal image is acquired, a confocal image acquisition operation is performed (Step S5), and then it is determined whether or not the objective lens has reached the vertical scanning end position (Step S6). If the objective lens has not reached the vertical scanning end position, the objective lens is moved in the vertical direction (step S7), and the steps after the acquisition of the low coherence image (step S3) are performed again. If the scanning end position has been reached, the image acquisition is terminated (step S8).

Note that low-coherence images need to be acquired at intervals sufficiently shorter than the coherence length (several μm to several tens of μm), but confocal images are set at intervals of about a fraction of the depth of focus. do it. In this way, the number of confocal image acquisitions can be reduced compared to a low-coherence image.
When the white light source 42 of the low coherence optical system 1B can be turned on / off at a relatively high speed such as a laser diode (LED), the white light source 42 itself is turned on / off without using a separate shutter or the like. May be.

Next, a procedure for obtaining the height of the target part (pixel) and the amount of emitted light from the data acquired in this way will be described with reference to FIG.
When the vertical scanning is completed, a pattern of height (z position) and interference output is created for each pixel of the acquired low coherence image data (step S11). An example of this pattern is “(4) white light interference waveform” in FIG.

Next, the envelope peak position, that is, the height of the portion corresponding to the target pixel is obtained from this pattern (step S12). As for the method of obtaining the envelope peak position of white light (low coherence) interference, the simplest method is to interpolate the data of each measurement height or take a moving average, but there are other methods. In order to perform highly accurate detection with as little data as possible, various methods have already been proposed (for example, Document 1 above).
Using the output data of the confocal optical system 1A at three or more heights adjacent to the calculated height, target pixel interference at five scanning image acquisition positions (heights) adjacent to the white light interference envelope peak The phase of the waveform is calculated (step S13), and then the emitted light intensity of the scanned image target pixel is calculated (step S14).

(4) In the cases of (1) to (3) above, the number of measurements can be reduced by obtaining the wavelength at the beginning of measurement by obtaining the wavelength at the beginning of measurement and reducing the number of measurements. In that case, the measurement is performed according to the following procedure.
First, the stage 28 is set at the vertical scanning start position.
Next, after starting vertical scanning, confocal data of one point or one line is acquired at four heights for pixels or lines in which a certain amount of light is first detected, and these confocal data are obtained by equation (1). The wavelength is calculated from

  When these confocal data are displayed on the screen, since the wavelength is constant over the entire screen, it is not necessary to take all the data on the screen, and in-plane scanning is stopped in the sub-scanning direction. Alternatively, only one line may be scanned, or only one point of data may be acquired without performing in-plane scanning at all.

  Thereafter, when taking the data of the entire screen, the data is taken three times by shifting the position by 1/8 of the wavelength (that is, 1/4 of the interference waveform S (z)) per cycle of the interference waveform, The peak position and output are obtained by a three-step method. Usually, it takes more time to acquire a confocal image by scanning than to acquire a white light image. Therefore, if the number of acquisition of confocal images decreases, the speed of measurement is expected to increase accordingly.

In addition, when a wavelength is calculated | required with each pixel of 1 line, this is good also as averaged as the measured value of a wavelength.
Further, since one cycle of the interference waveform is divided into four parts and moved, in addition to the three times when the confocal image is taken, one pixel or one line of data is acquired in the remaining one time, so that The wavelength may be calculated based on four times of data, and the wavelength may be corrected during the measurement. Thereby, the influence of the wavelength fluctuation during measurement can be reduced.

  In the above (1) to (4), when the branching ratio of the reference optical path is about 1/10, the amount of light entering the photodetector 35 is about 1/100. Even when the amount of light reaches about 10,000, it is possible to detect an interference pattern with reference light.

  As described above, according to the shape measuring apparatus 1 of the present embodiment, interference outputs at three or more measurement heights adjacent to the measurement target site on the sample S are detected, and the three or more measurement heights are detected. The operation of obtaining the peak position and amplitude of the interference output in the vicinity of is repeatedly performed for each measurement height, and the height of each measurement target portion of the sample S is determined based on the maximum value and the height of the obtained interference pattern. Since the amount of emitted light is calculated, both the horizontal resolution and the vertical resolution can be improved. Accordingly, it is possible to realize a three-dimensional shape measuring apparatus having a confocal resolution in the horizontal direction and having a resolution of the wavelength order or less regardless of the focal length of the objective lens in the vertical direction.

DESCRIPTION OF SYMBOLS 1 Shape measuring apparatus 1A Confocal optical system 1B Low coherence optical system 2 Control apparatus 11 1st illumination optical system 12 Coherent light source (1st light source)
21 scanning optical system 22 first scanning mirror 24 beam splitter (first branching means)
25, 27 Objective lens 26 Reference mirror (reflecting means)
28 Stage 29 Movable mechanism 31 First detection optical system 32 Beam splitter 35 Photo detector 41 Second illumination optical system (second optical system)
42 White light source (second light source)
51 Second detection optical system 52 Imaging lens 53 Image sensor S Sample

Claims (7)

A first optical source, a scanning optical system for converging the light beam emitted from the first light source on the surface of the sample while converging on the surface, and a detection optical system for detecting the optical beam interacting with the sample; A shape measuring apparatus comprising: a confocal optical system comprising: a variable means for changing the height of the sample up to the surface;
The confocal optical system is provided with a first branching unit for branching the light beam, and a reflecting unit for reflecting one light beam branched by the first branching unit and entering the detection optical system,
The operation of detecting the interference output at three or more measurement heights adjacent to the measurement target site on the sample and obtaining the peak position and amplitude of the interference output near the three or more measurement heights for each measurement height. A shape measuring apparatus, which is repeatedly performed every time, and calculates the height and the amount of emitted light of each measurement target portion of the sample based on the maximum value and the height of the amplitude of the obtained interference pattern.   The range for performing the phase analysis of the height is set to exceed one period of the interference pattern, and the relative phase from the start point of each measurement point that exceeds one period of the interference pattern from the start point of the range, The shape measuring apparatus according to claim 1, wherein phase analysis is performed corresponding to a range of one cycle.   3. The shape measuring apparatus according to claim 1, wherein the peak position and amplitude of the interference pattern are calculated using a phase shift (PSI) method. A second light source having lower coherence than the light source, and a second optical system for guiding a second light beam generated from the second light source onto the surface of the sample via the objective lens of the scanning optical system; And an imaging lens that forms an image of the second light flux that interacts with the sample, and an imaging device that is disposed on the imaging surface of the imaging lens, and the reflection means is shared with the confocal optical system. With coherence optics,
The optical distance from the reflecting means to the branching means is substantially equal to the optical distance from the branching means to the focal point of the objective lens, and is at the position of the low coherence image corresponding to the position to be measured for the height of the sample. From the height information, the height at which the intensity of the pixel with respect to the measurement site of the low coherence image reaches a peak is obtained, and the height around the peak height is obtained from the interference output data of the confocal optical system. The shape measuring apparatus according to claim 1, wherein an emitted light amount at a peak position in the confocal optical system is obtained.   5. The shape measuring apparatus according to claim 1, wherein the first branching unit has a branching ratio to the reflecting unit of 1/3 or less.   At the time of calibration or measurement start, four or more positions are selected from the height displacement for one period of the interference pattern obtained from the nominal wavelength of the first light source, and interference outputs are obtained, and these interferences are obtained. 6. The shape measuring apparatus according to claim 1, wherein an actual wavelength is obtained from the output data.   7. The shape measuring apparatus according to claim 6, wherein the interference output data acquired at the four or more positions is only data of one point or one line in the scanning plane on the surface of the sample.

Images