JP2006112872A - Compact angle sensor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a compact and highly sensitive optical angle sensor. <P>SOLUTION: Diffused light emergent from a spot light source 100 is passed through a polarization beam splitter 120 and a quarter wave plate 130 and modified into parallel light to irradiate a sample surface by a lens 170 combining a collimating lens and an objective lens. Reflected light is converged by the lens 170 and passed through the quarter wave plate 130. Light reflected at the polarization beam splitter 120 is detected by a light spot position detection element 160 arranged at a focal position in the optical angle sensor. Since it is possible to constitute the optical angle sensor of a few number of parts and acquire high sensitivity, the compact and highly sensitive optical angle sensor can be constituted. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an optical angle sensor.

  As the use of optical machines expands, the opportunity to require high-precision mirror surfaces of various shapes such as spherical surfaces, aspheric surfaces, cylinders and large flat surfaces is increasing. Large special optical elements used for X-ray optics, astronomical observation, etc. are large in the order of several hundreds of millimeters to meters, and high shape accuracy of 10 nm to submicron order is required. In recent years, semiconductor wafers and liquid crystal substrates have been increased in size and accuracy. In addition to high measurement accuracy, these shape measurements require measurement in a short time.

  Conventionally, the light wave interferometry of the surface measurement method has been used as a method of measuring the processed surface shape. This method has the advantage that the entire surface can be measured at once, but measurement may not be possible depending on the shape of the measurement sample, and it is very difficult to manufacture an interferometer with a diameter of 300 mm or more and to guarantee its accuracy. is there. On the other hand, the scanning method of measuring the shape by scanning the sensor has the advantage that it can flexibly cope with the enlargement and diversification of the shape to be measured, except for the problem of guidance accuracy. Furthermore, the possibility of on-machine measurement on machine tools can be expected.

  When an angle sensor is used for the scanning method, the local slope (local slope), which is a derivative of the shape, is measured by scanning the sensor or the measurement sample, and the shape is restored by integration. At this time, if an error is included in the output of the angle sensor, the error is accumulated by integration, and the shape cannot be obtained correctly. Therefore, high measurement accuracy is required for the angle sensor. In addition, the sensor needs to be small for scanning. An autocollimator is an example of an angle sensor that is currently widely used. Although the autocollimator has high measurement accuracy, the diameter of the light beam is as large as several tens of millimeters because it measures the tilt of the entire plane, not the local tilt. Also, since a CCD element is used, the response is slow and large, so it is not suitable for measurement by the scanning method.

  On the other hand, precision stages are frequently used in ultra-precision processing machines, semiconductor manufacturing / inspection equipment, precision measuring machines, office automation equipment, and the like. Since the movement error of these precision stages has a great influence on the performance of the machine, measurement of the movement error is important. Of the movement errors, positioning errors in the moving direction are measured by an interferometer or linear encoder. Moreover, the translation error (straightness) in the direction orthogonal to the movement error is measured based on a straight ruler. However, at present, the angle error around each axis (XYZ axis) is hardly measured. Conventional autocollimators are slow in response and cannot be used to measure dynamic angular errors. Furthermore, it is desirable that the sensor be as small as possible due to constraints on the work space and incorporation into the machine.

  Under such circumstances, the present invention proposes a highly sensitive, highly accurate and compact angle sensor.

  The angle sensor of the present invention uses an autocollimation method. The principle is shown in FIG. For simplification, FIG. 1 shows an example of detecting the rotation angle around the X axis.

The angle around the X axis with respect to the optical axis of the light incident on the collimator lens is defined as Δθ _X. Assuming that the focal length of the objective lens is f and the spot displacement in the Y direction on the focal plane of the lens is d _Y , the following relationship holds when Δθ _X is sufficiently small.

Here, a change in incident angle can be replaced with a change in distance. The value of d _Y depends on the angle of incidence, the beam if theta _X is constant is imaged on the same point on always the focal plane be incident anywhere in the objective lens. By using a detector (light spot position detector) that can detect the change in position of the light spot in the Y direction, it is possible to detect the angle θ _X around the X axis by detecting the change in d _Y. is there. The position d _Y in the Y direction and the X direction of the light spot, by using a detector capable of detecting simultaneously d _X, is possible to detect the rotation angle [Delta] [theta] _X and Y-axis rotation angle [Delta] [theta] _Y around the X-axis at the same time it can.

  As can be seen from FIG. 1, in the angle sensor based on the autocollimation method, the angle detection sensitivity / resolution is proportional to the position detection sensitivity / resolution of the detector that detects the position of the light spot and the focal length f of the objective lens. When a CCD element is used as a detector, the position detection resolution of the CCD element is about the same as the size of the CCD pixel. Because of the manufacturing principle, the pixel size of the CCD element is limited to the order of microns, so the position detection resolution is low. Therefore, in order to realize a highly sensitive and high resolution angle sensor, it is necessary to increase the focal length. However, if a lens with a long focal length is used, the sensor becomes large. In order to realize a highly sensitive and compact sensor, it is necessary to use a lens with a short focal length. For this purpose, it is necessary to use a light receiving element with high light spot position detection sensitivity.

  Therefore, the present invention proposes a method using an optical position detection element PSD (non-divided type) and a divided type PD, which can achieve high sensitivity even when using a lens with a short focal length. In addition, in the optical system of the angle sensor, a compact optical system has been devised that replaces two lenses with a single lens: a collimating lens for collimating the point light source and an objective lens for an autocollimation unit.

  Examples of the present invention will be described below.

First, the enhancement of sensitivity when PSD is used will be described.
FIG. 2 shows the principle of a two-dimensional semiconductor position detecting element (PSD). PSD is a non-divided element, has excellent linearity, and can continuously detect a two-dimensional position of the center of gravity of a spot in the XY direction with one element. There is also a feature that is not affected by the light intensity distribution. The width of the light receiving surface of the two-dimensional PSD XY directions together To L _P, when the two-dimensional position of the light spot detected thereby d _X, and d _Y, d _X, d _Y is PSD optical current output I _X1 , I _X2 , I _Y1 , I _Y2 . From the two-dimensional PSD X, Y output x _{out_PSD} , y _{out_PSD,} it can be obtained as follows.

As can be seen from the above equations, the PSD position detection sensitivity defined by x _{out_PSD} / d _X and x _{out_PSD} / θ _Y is mainly determined by the width of the light receiving surface. Since the position detection sensitivity is inversely proportional to the width of the light receiving surface, a shorter light receiving surface width is more advantageous for higher sensitivity. In other words, the angle sensor can be made highly sensitive by reducing the PSD light receiving surface width. In addition, it is necessary to determine the PSD's light-receiving surface width from the required sensor resolution and sensitivity. For example, when the focal length of the objective lens is 40 mm, the spot movement amount d _X (or d _Y ) on the PSD corresponding to the inclination θ _Y (or θ _X ) of the sample surface of 0.01 seconds is 4 nm. If the required angular resolution of the sensor is 0.01 seconds and the dynamic range of the sensor (ratio of the measurement range to the resolution) is 10,000, the required light-receiving surface width is up to about 40 μm. Although the figure shows a two-dimensional PSD, the same can be said for a one-dimensional PSD.

Next, a description will be given of the enhancement of sensitivity when a split PD is used.
FIG. 3 shows the principle of the quadrant PD that detects the light spot position. The quadrant PD can detect the beam spot position by measuring the current from each element in proportion to the home of the beam when the beam is incident on each light receiving element.

ここで全素子からの出力でわることで光量の変化の影響を補正している。 Here, the outputs from the elements A, B, C, and D are I _A , I _B , I _C , and I _D , respectively, and the X direction output X _{out_PD} and the Y direction output Y _{OUT_PD} are defined as follows.

Here, the influence of the change in the amount of light is corrected by the output from all the elements.

  The spot needs to be incident on all the elements of the quadrant PD. Since the element is sufficiently larger than the spot size, the measurement range is determined by the spot size.

となる。それぞれのPD上の光スポットの面積はPD素子の出力電流と比例しているので，式(6)(7)を式(4)に代入すると

となる。 Next, the relationship between the position detection sensitivity of the light spot by the divided PD and the light spot size on the PD will be considered. FIG. 4 shows an example of detecting the d _X. As shown in FIG. 4, the spot center is displaced by d _{X in the} X direction with a diameter of 2w. When d _X is small, the portion that has changed since the spot was at the center can be approximated as a rectangle. The area where the light hits the two sets of left and right PD elements on the X axis

It becomes. Since the area of the light spot on each PD is proportional to the output current of the PD element, substituting Equations (6) and (7) into Equation (4)

It becomes.

となる。ここで，式（１），（８），（９）を用いて角度の検出感度を計算すると

となり整理すると

となる。 Further, as shown in FIG. 5, based on the theory of light diffraction, if the focal length of the objective lens is f, the wavelength of the light beam is λ, and the beam diameter incident on the lens is D, the spot size 2w on the focal plane is

It becomes. Here, if the angle detection sensitivity is calculated using equations (1), (8), and (9),

When it becomes

It becomes.

  In other words, sensitivity does not depend on the focal length. It can be seen that this depends only on the wavelength of the light incident on the objective lens and the incident beam diameter. That is, the angle sensor can be made highly sensitive by shortening the wavelength of light to be used and increasing the incident beam diameter of the objective lens. This makes it possible to select a lens without worrying about a decrease in sensitivity due to the focal length.

The downsizing of the angle sensor will be described.
First, FIG. 6 shows a layout example of the optical system of the conventional two-dimensional angle sensor. A linear P-polarized light beam emitted from a point light source such as a laser diode (LD) is collimated by a collimating lens, and passes through a polarizing beam splitter (PBS) and a quarter-wave plate (1 / 4λ) to be circularly polarized. The measurement sample is irradiated. The reflected light with the angle information of the sample surface passes through the quarter-wave plate again, and the polarization state changes to S-polarized light. The light is reflected by the polarization beam splitter, and then enters an autocollimation unit comprising an objective lens and a light spot position detecting element placed at the focal plane position. The angle of the sample surface can be obtained by measuring the displacement of the light spot with the light spot position detecting element. As the light spot position detecting element here, PSD, split type PD, CCD element, CMOS element or the like is used.

  FIG. 7 shows an optical system capable of realizing a small angle sensor. In order to achieve a compact design, the collimating lens for collimating the light from the point light source and the objective lens for the auto-collimation unit are performed by a single lens. Up to now, two required lenses can be combined into a single lens, so the number of parts can be reduced and the overall size can be reduced. In this optical system, the linear P-polarized light beam emitted from a point light source such as a laser diode (LD) is converted into parallel light by a lens, passes through a polarization beam splitter and a quarter-wave plate, and becomes circularly polarized light. The sample is irradiated. Reflected light with the angle information of the sample surface enters the same lens again, passes through the quarter-wave plate, and becomes S-polarized light. The light is reflected by the polarization beam splitter and focused on the light spot position detecting element placed at the focal plane position of the lens. The angle of the sample surface can be obtained by measuring the displacement of the light spot with the light spot position detecting element. As the light spot position detection element here, a PSD, a split type PD, a CCD element, or a CMOS area sensor CMOS area sensor (for example, profile sensor S9132 manufactured by Hamamatsu Photonics Co., Ltd.) is used.

At this time, if the length of one side of the PBS is L, the thickness of the 1 / 4λ plate is T1, the thickness of the lens is T2, and the focal length of the lens is f, the lens is not focused in PBS. ,
f ≧ L + T1 + T2 / 2
Must. Since T1> 0 and T2> 0,
f> L
Must. Also, if the diameter of the light beam collimated by the lens is D, and the divergence angle of the light emitted from the LD is φ,
D = 2 · f · tanφ
It becomes. In order that the light emitted from the LD does not leak to the side of the PBS, L ≧ D
Must. Therefore, if f, L, and φ are design parameters,
f> L ≧ 2 · f · tanφ
It is necessary to satisfy the relationship.

  Based on the optical system shown in FIG. 7, as shown in FIG. 8, when the LD, PBS, 1 / 4λ plate, and lens are directly arranged on the semiconductor silicon substrate on which the light spot position detecting element is built, , The sensor can be realized in a compact manner.

  As shown in FIGS. 9 and 10, by reducing the size of PBS, a lens having a short focal length can be used, and a small sensor can be realized.

  According to the present invention, an optical angle sensor capable of measuring with a small size and high accuracy can be constructed at low cost. In addition to using the angle sensor alone, the angle of the stage can be incorporated into a stage of various measuring instruments, machine tools, OA equipment, etc. It is possible to measure the movement error in real time and correct the error.

It is a figure which shows the principle of the autocollimation method. It is a figure explaining the light spot position detection principle by PSD. It is a figure explaining the light spot position detection principle by split type PD. Relationship between spot detection sensitivity and spot diameter by split PD. The figure explaining the spot diameter on an objective-lens focal plane. The example of the optical system of the conventional angle sensor. The optical system of the small angle sensor of this invention. An example of a small angle sensor in which components on a semiconductor silicon substrate are directly arranged. Example of a small sensor using a lens with a short focal length (Example 2). Example of a small sensor using a lens with a short focal length (Example 3).

Explanation of symbols

100 point light source 101 laser diode 102 LD case 110 collimating lens 120 PBS (polarization beam splitter)
130 1/4 wavelength plate 140 Sample surface 150 Objective lens 160 Light spot position detection element 170 Collimator / Objective lens 180 Spacer 200 Light beam 300 Silicon substrate

Claims (2)

  After the diffused light emitted from the point light source 1 passes through a polarizing beam splitter and a quarter-wave plate, it is converted into a parallel light beam by a lens serving as a collimating lens and an objective lens, and irradiated on the sample surface. An optical angle sensor characterized in that light converged by a lens that also serves as an objective lens, passes through a quarter-wave plate, and is reflected by a polarizing beam splitter is detected by a light spot position detecting element placed at a focal position. 2. The optical angle sensor according to claim 1, wherein the divergence angle of the diffused light is φ, the length of one side of the polarizing beam splitter is L, and the focal length of the lens serving as the collimating lens and the objective lens is f. An optical angle sensor satisfying the relationship of ≧ 2 · f · tanφ.

Images