A scanning probe microscope and a measuring method using the same

The present invention relates to a scanning probe microscope used for applications such as high-precision measurement of sample shape.

With the increasing integration of semiconductor circuits, the importance of inspection measurement technology and defect analysis technology in the semiconductor manufacturing process is increasing as circuit patterns continue to be miniaturized. As the recording density increases, the hard disk device also measures the fine structure and flatness of the magnetic poles of the recording / playback head, the surface roughness of the recording medium, and the three-dimensional shape of the magnetic stripe or dot structure to further increase the recording density. Is becoming more important. A scanning probe microscope (hereinafter referred to as SPM; Scanning Probe Microscope in the present specification), which is most suitable for such an application, scans a probe while bringing the tip of a minute probe close to or in contact with the sample surface. Therefore, it is widely known as a method for measuring the shape of the sample surface on the atomic order.

In surface shape measurement using SPM, the inspection area is limited to a narrow area of several hundreds of micrometers or less, but when measuring a minute area of atomic order, a narrow field of view of several tens to several hundreds of nanometers is used. Is required to be measured with an accuracy less than the atomic order. At this time, high positioning accuracy is required for the mechanical unit for scanning the probe. On the other hand, in order to find the measurement area, it is required to observe a wide range of several tens of micrometers at high speed. There is also a need to measure the unevenness of the sample surface in a wide range of about several hundred micrometers at high speed.

In addition, SPM has the advantage of being able to measure the three-dimensional shape of the sample surface with a high resolution of about 0.1 nanometer, but it takes time to position the measurement points on the sample surface and to perform measurement operations, and sufficient measurement throughput cannot be obtained. Therefore, it is not used in-line (during the manufacturing process) in device manufacturing lines such as semiconductors and hard disk devices, but is mainly used for offline defect analysis. However, if abnormalities in various process equipment can be immediately detected from the measurement results by SPM and fed back to the processing conditions of various process equipment, the production of defective products can be minimized and the manufacturing yield of the production line can be improved. It is expected that inline SPM will be realized because it can be done. That is, when realizing in-line SPM, how many measurement points the SPM can process (measure) per unit time is an important factor, and in the current manufacturing line, processing of 20 seconds or less per point is an important factor. Time is required. This is equivalent to 30 WPH (Wafer Per Hour) or more when converted to measured throughput.

A piezoelectric element is usually used as an actuator of a mechanism for positioning an SPM probe on a sample with high accuracy. For example, in Patent Document 1, the three axes X, Y, and Z are each composed of parallel flat plates, and these are driven by piezoelectric elements, and at the same time, the position of the probe is measured with a displacement meter to determine the position of the probe. It is described that high-precision SPM is realized by controlling. As another probe driving mechanism for improving the positioning accuracy of the probe, there is a three-dimensional microscanning mechanism disclosed in Patent Document 2. This is a 3-axis stage with 3 voices forming an XZ stage (combined with the X stage and Z stage) connected to the Y stage by an elastic member in the Y stage connected to the outer frame by an elastic member. It is a mechanism driven by a coil motor.

All stages are integrally formed of the same member, and the driving force of the voice coil motor is transmitted to each stage via a spindle. Regardless of the displacement of each stage, each spindle is always pressed parallel to the operating direction of each stage. For example, when only the Y stage operates, all the elastic members connecting the outer frame and the Y stage are elastically deformed evenly, so that no unnecessary force is applied to the operating axes other than the Y axis. As described above, a probe scanning mechanism capable of controlling the positioning of the probe independently on three axes and with high accuracy has been realized. Further, Patent Document 3 describes a method of improving the positioning resolution of the stage by using a piezoelectric element in which two types of piezoelectric elements for fine movement and coarse movement are connected.

Further, Patent Document 4 discloses a configuration of SPM for improving measurement throughput. This is because the SPM detects the sample surface position with an approach sensor consisting of an objective lens, a laser diode, and a photodiode placed directly above the probe, and brings the sample surface closer to the probe tip position at high speed. It shortens the time to start the measurement operation and improves the measurement throughput of SPM. According to the SPM having the configuration shown in Patent Document 3, since the objective lens is arranged directly above the probe contact position on the sample, the observation optical system is used on the sample. After the measurement position is determined, the measurement operation can be performed without moving the sample position, and the SPM measurement throughput can be improved.

<p num="0008"><patcit num="1"><text>Japanese Unexamined Patent Publication No. 2004-303991</text></patcit><patcit num="2"><text>Japanese Patent No. 3544453</text></patcit><patcit num="3"><text>Japanese Patent Application Laid-Open No. 2005-347484</text></patcit><patcit num="4"><text>Japanese Unexamined Patent Publication No. 2004-125540</text></patcit></p>

<p num="0009"> With conventional SPM, it is difficult to achieve both rapid scanning of a wide range of probes and accurate scanning of a small range of probes with high resolution, and measurement of a narrow range is performed at the expense of operability. We had no choice but to configure an SPM that only performed, or to be able to measure a wide range at the expense of accurate and high resolution measurements over a narrow range.</p><p num="0010"> An object of the present invention is a scanning probe that solves the above-mentioned problems of the prior art and enables rapid scanning of a wide range of probes and accurate scanning of a small range of probes with high resolution. It is to provide a microscope and a measurement method using the microscope.</p>

<p num="0011"> In order to solve the above problems, in the present invention, the cantilever and the tip of the cantilever are fixed. A probe unit equipped with a fixed probe and a probe unit are driven to bring the probe close to or in contact with the sample surface. A probe unit driving means that scans the sample surface while repeating the process and separating the sample, and the probe unit. Detects the amount of displacement of the probe part when the probe is in close proximity to or in contact with the sample surface by the driving means. The shape of the sample surface based on the displacement amount detecting means and the displacement amount of the probe portion detected by the displacement amount detecting means. In a scanning probe microscope equipped with a signal processing means for obtaining state information, a probe<u style="single">Department</u>The drive means has a drive unit that independently drives the probe in three orthogonal axial directions, and the drive speed in at least one axial direction while the probe unit is being driven by the drive unit.<u style="single">And response speed</u>It is configured with a drive speed switching unit that switches between. Further, in order to achieve the above object, in the present invention, the cantilever and the tip of the cantilever are attached. A probe unit equipped with a probe fixed close to the sample is driven to bring the probe close to or in contact with the sample surface. The probe unit that scans the sample surface while repeatedly separating it from the sample surface. Scanning type that detects the displacement amount of the sample surface and obtains the shape information of the sample surface based on the detected displacement amount of the probe unit. When scanning the sample surface by driving the probe in the measurement method using a probe microscope At least one axial drive speed<u style="single">And response speed</u>Was made to scan while switching while driving.</p>

<p num="0012"> According to the present invention, in SPM, driving by two types of actuators, short stroke and long stroke, is combined, and the former has high resolution and high speed response, and the latter has long stroke and low speed response, so that high resolution and long stroke can be achieved. Achieves probe positioning and scanning. By incorporating a responsiveness switching mechanism in the driver so that the responsiveness of the driver for long strokes is increased only when the voltage change is large, both high speed during long stroke drive and high stability during positioning are achieved. Let me. This provides a probe microscope capable of scanning a long stroke on the order of several tens of micrometers to several hundreds of micrometers and realizing probe position accuracy on the order of sub-nano to picometer.</p><p num="0013"> In addition, since long-stroke drive can be performed at high speed, if this is applied to the Z-axis, it is possible to speed up the approach of the probe to the sample before the start of measurement. In addition, when applied to the XY axes, it is possible to scan a wide range at high speed to quickly find the pattern to be measured and scan the location with high accuracy. These can improve the throughput of SPM.</p><p num="0014"> Further, if the SPM according to the present invention is applied to a manufacturing process of a semiconductor, a hard disk, or the like, it is possible to optimize the processing conditions of the manufacturing apparatus based on the measurement result by the SPM, so that the yield of the device manufacturing process is improved. To do.</p>

<figref num="1">It is a block diagram which shows the structure of the whole apparatus of a scanning probe microscope.</figref><figref num="2A">It is a top view which shows the schematic structure of the probe drive part of a scanning probe microscope.</figref><figref num="2B">It is sectional drawing of the front which shows the schematic structure of the probe drive part of a scanning probe microscope.</figref><figref num="2C">It is a side view which shows the schematic structure of the probe drive part of a scanning probe microscope.</figref><figref num="3">It is a front view which shows the schematic structure of the probe deflection detection optical system.</figref><figref num="4A">It is a top view which shows the structure of the main part of the probe drive part and the driver circuit part of a scanning probe microscope.</figref><figref num="4B">It is a front view which shows the schematic structure of the probe drive part of a scanning probe microscope.</figref><figref num="5A">It is a graph which shows the time change of the coarse displacement of a piezoelectric element driver in a conventional scanning probe microscope, and shows the case where the time constant is small.</figref><figref num="5B">It is a graph which shows the time change of the coarse displacement of a piezoelectric element driver in a scanning probe microscope, and shows the case where the time constant is large.</figref><figref num="6A">The figure which shows the circuit structure of the piezoelectric element driver for coarse motion shows the structure which switches the output resistance Rs for high-speed drive by the responsiveness switching signal from the outside.</figref><figref num="6B">The figure which shows the circuit structure of the piezoelectric element driver for rough motion shows the structure which inserts a pair of diodes in opposite directions and automatically switches the output resistance Rs for high-speed drive.</figref><figref num="6C">The figure which shows the circuit structure of the piezoelectric element driver for rough motion shows the structure which inserts a pair of diodes in opposite directions and divides and inserts a resistor to automatically switch a high-speed drive output resistor Rs.</figref><figref num="6D">It is a figure which shows the circuit structure of the piezoelectric element driver for fine movement.</figref><figref num="6E">It is a graph which shows the time change of the coarse displacement with respect to the step-like input.</figref><figref num="6F">It is a graph which shows the time change of the fine movement displacement with respect to a step-like input.</figref><figref num="7">It is a figure which shows the Example which further improves the resolution of the piezoelectric element driver for fine movement.</figref><figref num="8A">An open control configuration is used in a control block diagram using a responsive switching amplifier and a slightly coarse-moving piezoelectric element.</figref><figref num="8B">It is a figure which shows the example which uses the signal which reduced the noise by filtering the position detection signal of a stage in the control block diagram which used the responsiveness switching amplifier and the fine coarse piezoelectric element.</figref><figref num="8C">It is a figure which shows the example of the configuration which provides the dead zone in the control block diagram which used the responsiveness switching amplifier and the fine coarse-moving piezoelectric element, and controls.</figref><figref num="9A">It is a graph which shows the state of the time change of the coarse displacement when the command position changes greatly in a step shape.</figref><figref num="9B">It is a graph which shows the state of the time change of the fine movement displacement when the command position changes greatly in a step shape.</figref><figref num="9C">It is a graph which shows the state of the time change of the total displacement when the command position changes greatly in a step shape.</figref><figref num="10A">It is a graph which shows the state of the time change of the coarse displacement when the command position changes in a step shape smaller than the level of a dead zone.</figref><figref num="10B">It is a graph which shows the state of the time change of the fine movement displacement when the command position changes in a step shape smaller than the level of a dead zone.</figref><figref num="10C">It is a graph which shows the state of the time change of the total displacement when the command position changes in a step shape smaller than the level of a dead zone.</figref><figref num="11A">It is a circuit diagram which shows the structure which switches and drives a coarse-moving piezoelectric element and a fine-moving piezoelectric element by an external input using only one amplifier.</figref><figref num="11B">It is a circuit diagram which shows the structure which automatically switches and drives a coarse-moving piezoelectric element and a fine-moving piezoelectric element by using only one amplifier.</figref><figref num="12A">It is a top view which shows another example of the mechanism which drives a probe.</figref><figref num="12B">It is a front view which shows another example of the mechanism which drives a probe.</figref><figref num="13">It is a top view which shows another example of the mechanism which drives a probe.</figref><figref num="14">It is a top view which shows another example of the mechanism which drives a probe.</figref><figref num="15A">It is a block diagram which shows a part of the semiconductor manufacturing process using a scanning probe microscope (SPM).</figref><figref num="15B">It is a flow chart which shows the process flow at the time of measuring a sample using a scanning probe microscope (SPM) in a semiconductor manufacturing process.</figref><figref num="16A">It is a front view of the probe which is measuring the sample surface.</figref><figref num="16B">It is a graph which shows the time change of the height of the sample surface and the tip of the probe when the sample surface is measured by the probe.</figref><figref num="17A">It is a front view of the probe which shows another example of measuring a sample surface.</figref><figref num="17B">It is a graph which shows the time change of the height of a sample surface and a probe tip which shows another example when the sample surface is measured by a probe.</figref><figref num="18A">It is a top view of the sample which shows the state which detected the measurement area by scanning the probe greatly by coarse movement.</figref><figref num="18B">It is a front view of the probe which shows the state of measuring the measurement area.</figref><figref num="19">18 is a flow chart of processing corresponding to the operations of A and B.</figref><figref num="20">It is a perspective view of the optical interference displacement sensor.</figref><figref num="21A">It is a perspective view and a partially enlarged view which show the structure of the phase shift element used for an optical interference displacement sensor.</figref><figref num="21B">It is a perspective view of the diffraction polarization lattice used as a reference mirror of an optical interference displacement sensor.</figref><figref num="21C">It is a perspective view of the photonic crystal used as a reference mirror of an optical interference displacement sensor, and the perspective view of a photonic crystal used as a 1/4 wave plate.</figref>

As a first embodiment of the present invention, first, the configuration of the SPM which is the basis of the present invention will be described with reference to FIGS. 1 to 3. In FIG. 1, 103 is a sample to be measured, and 104 is a sample stage for vacuum-adsorbing the sample 103 and moving the sample 103 in the X, Y, Z directions and the rotation direction in the XY plane. The operation of the sample stage is controlled by the stage control unit 111. The probe 102 is held by the probe drive mechanism 101 via the probe holder 115.

The probe drive mechanism 101 precisely positions the probe 102 on the sample 103 in the X, Y, and Z directions. The probe 102 is made of a silicon material or the like, and its tip diameter is processed to 10 nanometers or less by etching and a slightly focused ion beam. Alternatively, carbon nanotubes having a diameter of about 10 nm may be formed at the tip. The probe 102 is composed of a cantilever and a probe formed at its tip, but in the present specification, the cantilever and the probe are collectively referred to as a probe. An observation optical system lens barrel 105 provided with an objective lens 106 is arranged on the probe drive mechanism 101.

The observation optical system 105 has an imaging camera inside, and the optical image of the surface of the sample 103 magnified by the objective lens 106 is displayed on the TV camera 107 via the optical image processing unit 108. The observation optical system 105 and the objective lens 106 include a focus axis that moves up and down in the Z direction by a moving mechanism (not shown). A small piezoelectric element may be incorporated in the probe holder 115 so that the held probe 102 can be vibrated with an amplitude on the order of several nanometers to several tens of nanometers.

2A to 2C are explanatory views showing the structure of the probe drive mechanism 101 shown in FIG. 1, FIG. 2A is an XY plan view of the probe drive mechanism 101, and FIG. 2B is A-A of the probe drive mechanism 101. 'A cross-sectional view taken along the arrow, FIG. 2C is a view showing a YZ plan view of the probe drive mechanism 101, respectively. In the probe drive mechanism 101, the holders 201 and 202 and the Y stage 203 are integrally formed in the same plane via the elastic deformation portions 204a, 204b, 204c, and 204d, and the X stage is further formed in the same plane of the Y stage 203. The 207 is orthogonal to the Y stage 203, and has a structure integrally formed via elastically deformed portions 208a, 208b, 208c, and 208d. The X stage 207 is provided with through holes 211 for penetrating the objective lens 106. Laminated piezoelectric elements (hereinafter simply referred to as piezoelectric elements in this embodiment) 205 and 206 are adhered between the holders 201 and 202 and the Y stage 203, and the Y stage 203 has equal amounts of the piezoelectric elements 205 and 206. It is driven in the Y-axis direction by expanding and contracting at the same time. The drive mechanism composed of the piezoelectric element 205 and the elastically deformed parts 204a and 204b and the pair of drive mechanisms composed of the piezoelectric element 206 and the elastically deformed parts 204c and 204d are located at the center of the field of view 212 of the objective lens (probe 102). It is placed at a target position centered on the tip position).

A general piezoelectric element (piezoelectric ceramic element) is an element whose length changes by applying a DC voltage. A laminated piezoelectric element in which an electrode and a thin piezoelectric element are laminated is often used in order to obtain a large displacement with a small voltage. An example of a laminated piezoelectric element is one that has a length of 40 mm and is stretched by 20 micrometers by applying a voltage of 100 V. In the case of this piezoelectric element, if the voltage noise is about 5 mV, the resolution is the value obtained by multiplying the movable distance by the ratio of the noise and the maximum applied voltage, and the resolution is 1 nanometer. The configuration for achieving sub-nanometer-order resolution will be described later. Piezoelectric elements 209 and 210 are adhered between the Y stage 203 and the X stage 207, and the X stage 207 is driven in the X-axis direction by simultaneously expanding and contracting the piezoelectric elements 209 and 210 by equal amounts. The drive mechanism composed of the piezoelectric element 209 and the elastically deformed parts 208a and 208b and the pair of drive mechanisms composed of the piezoelectric element 210 and the elastically deformed parts 208c and 208d are located at the center of the field of view 212 of the objective lens (probe 102). It is placed at a target position centered on the tip position). The maximum movable distance of the piezoelectric elements 209 and 210 is 20 micrometers, and the movable resolution is 1 nanometer.

On the bottom surface of the X stage 207, the Z-axis mechanism unit 213 is attached so as to be orthogonal to the movable planes of the Y stage 203 and the X stage 207. In the Z-axis mechanism portion 213, the fixing portions 218 and 219 and the Z stage 214 are integrally formed in the same plane via the elastically deformed portions 215a, 215b, 215c and 215d. Piezoelectric elements 216 and 217 are adhered between the fixed portions 218 and 219 and the Z stage 214, and the Z stage 214 is driven in the Z-axis direction by simultaneously expanding and contracting the piezoelectric elements 216 and 217 by an equal amount. .. The drive mechanism composed of the piezoelectric element 216 and the elastically deformed parts 215a and 215b and the pair of drive mechanisms composed of the piezoelectric element 217 and the elastically deformed parts 215c and 215d are in the XZ plane with respect to the optical axis 212'of the objective lens. It is placed in the target position above. The maximum movable distance of the piezoelectric elements 216 and 217 is 10 micrometers, and the movable resolution is 1 nanometer. The probe 102 is attached to the Z stage 214 via the probe holder 115, and the tip position of the probe 102 coincides with the field center position 212 of the objective lens.

As described above, in the probe drive mechanism 101 according to the present invention, the operations of the X stage 207, the Y stage 203, and the Z214 stage for three-dimensionally driving the probe 102 do not interfere with each other and are independent of each other. It is possible to operate. Further, for example, in the Y stage 203, a set of stage drive mechanism units composed of two elastically deformed parts 204a and 204b arranged on the extension line of the expansion and contraction shaft of the piezoelectric element 205 with the piezoelectric element 205 sandwiched between them is Y. Since a pair (piezoelectric element 206, elastic deformation part 204c, 204d) is arranged on both the left and right sides of the stage 203, the piezoelectric elements 205 and 206 expand and contract by equal amounts, respectively, so that the elastic deformation parts 204a, 204b, 204c , 204d can all be deformed evenly. As a result, it is possible to eliminate the Abbe error of the Y stage 203 and significantly improve the straightness of the Y stage 203 as compared with the conventional case. It is obvious that this operating principle is exactly the same for the X stage 207 and the Z stage 214. The operations of the X stage 207, the Y stage 203, and the Z stage 214 of the probe drive mechanism 101 are controlled by the probe drive control unit 110.

Note that the laminated piezoelectric element may have individual differences in expansion and contraction displacement when a voltage is applied. Further, even if the same piezoelectric element is used, a hysteresis characteristic exists between the applied potential and the displacement. In this case, different hysteresis characteristics may be measured in advance for each piezoelectric element, and the applied voltage for achieving a desired displacement may be adjusted and applied for each piezoelectric element. Further, in this embodiment, a piezoelectric element is used to operate the X stage 207, Y stage 203, and Z stage 214 of the probe drive mechanism 101, but the power source of each stage is not limited to the piezoelectric element. , A linear actuator having sufficient accuracy and generating force necessary for positioning the probe 102 may be used. As the constituent material of the probe drive mechanism 101, a material such as an aluminum alloy or titanium having a large ratio of rigidity to specific gravity, or an iron-nickel alloy having a low coefficient of thermal expansion (linear expansion coefficient) is used.

The observation optical system lens barrel 105 and the objective lens 106 arranged above the probe drive mechanism 101 can be moved up and down in the Z-axis direction by a movement mechanism (not shown), and the objective lens 106 is attached to the probe drive mechanism 101. It is inserted into the through hole 211 provided in the X stage 207 so as not to come into contact with it. According to the configuration of the probe drive mechanism 101 of the present invention, since there is no mechanism for scanning the probe 102, the probe 102 can be directly observed with the objective lens 106, and at the same time, the surface of the sample 103 has a high resolution. It is possible to observe at. For example, if the aperture ratio of the objective lens 106 is 0.7 and the working distance is 6 mm, the pattern on the sample 103 can be clearly observed under the condition that the resolution is 1 micrometer or less. The field of view of the objective lens 106 is obtained by lowering the objective lens 106 and the sample stage 104 by the same amount (for example, 1 mm) so that the objective lens 106 does not come into contact with the probe 102 while the position of the probe 102 is fixed. It is also possible to observe the pattern on the sample 103 directly under the probe 102 without being affected by the presence of the probe 102 arranged inside. This utilizes an optical phenomenon obtained under the condition that the aperture ratio of the objective lens 106 is large and the probe 102 occupies only a part of the field of view of the objective lens 106.

FIG. 3 is an explanatory view showing the configuration of the probe deflection detection unit for detecting the contact between the probe 102 and the surface of the sample 103, and is an XZ plan view of the probe drive mechanism 101. Reference numeral 301 denotes a laser diode having an oscillation wavelength of 600 nanometers and an oscillation output of 0.1 milliwatt. The laser light oscillated from the laser diode 301 is shaped into parallel light by the collimating lens 302 and attached to the holder 201. It is folded back in the Y-axis direction (not shown), reflected again in the X-axis direction by the mirror 304 (not shown) attached to the Y stage 203, and irradiated to the back surface of the probe 102 via the mirrors 305 and 306. Laser. The laser beam reflected from the back of the probe 102 is reflected by the mirrors 307 and 308, folded back in the Y-axis direction by the mirror 309 (not shown) attached to the Y stage 203, and attached to the holder 202. At 310 (not shown), it is reflected again in the X-axis direction and received by the receiver 311. Here, the laser diode 301 is fixed to the holder 201 of the probe drive mechanism 101, the receiver 311 is fixed to the holder 202, and the mirrors 305, 306, 307, and 308 are fixed to the Z stage 214 by a jig (not shown). Regardless of the position of 102, the amount of deflection of the probe can be detected as a change in the laser irradiation position on the laser receiving surface of the receiver receiver 311.

As the receiver 311, a PSD (position sensitive device), an image sensor, a 2-split or 4-split photodiode, or the like can be used. When the probe 102 and the sample 103 come into contact with each other and the probe 102 is bent, the laser irradiation position moves in the Y-axis direction on the light receiving surface of the receiver 311. The receiver 311 converts this change in the laser irradiation position into a voltage signal, and the probe deflection detection unit 109 detects the contact between the probe 102 and the surface of the sample 103. When the probe 102 is vibrated by the probe holder 115 or the like, the probe deflection detection unit is configured to detect the amplitude and phase of the vibration of the probe deflection signal due to the deflection of the probe 102. Thereby, the force acting between the probe 102 and the sample 103 may be detected. That is, when the tip of the vibrating probe approaches the sample, the vibration state such as the vibration amplitude, the phase of vibration with respect to the vibration signal, and the vibration frequency changes due to the force acting between the tip of the probe and the sample. By detecting, the force can be measured.

Next, using FIGS. 4A and 4B, the configuration of a wide range, high speed, and high precision SPM using the coarse and fine movement mechanism, the responsiveness switching mechanism, and the displacement meter according to the present invention will be described.

First, the probe position measurement in the Z direction will be described. A part of the probe holder 115 faces the Z-axis capacitance sensor 224 fixed to at least one of the holders 201'and 202' or the bottom surface of the X stage 207'by a mechanism (not shown). It is configured. The distance between the probe holder 115 and the capacitance sensor 224 on the facing surface is 20 micrometers. The Z-axis capacitance sensor 224 used here can measure the distance from the probe holder 115 with a resolution of 0.1 nanometer, and measures the amount of movement of the Z stage 214'. The probe holder 115 and the Z-axis capacitance sensor 224 are arranged on the Z-axis that overlaps the optical axis 212'of the objective lens on the XZ plane.

With this arrangement, it is possible to measure the stage displacement (displacement of the tip position of the probe 102) at the center position 212 of the field of view of the objective lens, and even if there is a yawing error in the operation of the Z stage 214', there is almost no error in Abbe. It is a configuration that does not occur. The probe holder 115 is made of a metal material that constitutes the Z stage 214'and a metal material that provides electrical continuity, and the surface facing the Z-axis capacitance sensor 224 is subjected to precision grinding. There is. By using a more accurate displacement sensor with a resolution of around 10 picometers, such as a laser interference displacement meter as described later, instead of the capacitance sensor, SPM measurement with even higher accuracy and resolution is realized. it can.

As for the configuration on the sample 103 side, in the examples of FIGS. 4A and 4B, the coarse Z stage 403 is mounted on the air slider 404, the sample side XY stage 401 is mounted on the air slider 404, and the sample 103 is mounted on the sample 103 side. ing. Further, the air slider 404 and the coarse Z stage 403 are connected to the coarse XY stage 402 by an elastic plate 406, and the slider 404 is placed on the surface plate 405. The coarse XY stage 402 is fixed to the surface plate 405 by a structure (not shown). The selection of the observation position on the sample 103 is realized by moving the air slider 404 in the XY plane by the coarse XY stage 402 via the elastic plate 406. Compared to the normal method of stacking the X stage, Y stage, and Z stage, the height is smaller and the rigidity at rest can be kept high, so the vibration of the sample 103 can be almost eliminated. This is because it is possible and is most suitable for scanning probe microscopes. When moving the coarse XY stage 402, air may be blown between the air slider 404 and the surface plate 405 to reduce the frictional force, or the surface plate 405 may be driven without blowing air into the air slider 404. You may slide against it to move. The latter, especially when measuring a large area on the sample, such as hundreds of micrometers to tens of millimeters, with the probe 102, especially when using a flat surface plate 405 and sliding the air slider 404 over it, the probe 102. When measuring the surface shape of the sample 103 by means of, it is an effective moving method because there is no gap due to air.

Furthermore, when the probe 102 measures an intermediate region of several tens of micrometers to several hundreds of micrometers on the sample 103, the sample-side XY stage 410 is driven. This is a stage that is driven by a piezoelectric element and uses an elastic guide and has a structure similar to that of the stage on the probe side, but it does not have a Z-axis 214'correspondence to the stage on the probe side. Since the design can be made with an emphasis on the movable range rather than the rigidity required for high-speed scanning of the probe, the stage has a wide movable range of several hundred micrometers.

Since it is difficult to scan a wide range with high flatness, the vertical movement during scanning of the sample stage 401 is measured and scanned using a displacement sensor 410 such as a capacitance sensor or a laser interference displacement meter. By correcting the height data at each point of the measurement result of the type probe microscope, a measuring station with higher flatness can be realized. For example, if the scanning flatness of the stage 401 is 10 nanometers and the accuracy of the displacement sensor 401 is 0.1 nanometers, the measurement accuracy of the flatness of the measurement result by the scanning probe microscope by the correction described above Can be improved from 10 nanometers to 0.1 nanometers.

In the above explanation, the measurement result was corrected by the displacement meter 410, but the probe 102 is fed back directly to the Z stage 214'to follow the vertical movement of the sample stage 401 on the upper surface of the sample 103. May correct the height of Next, the probe position measurement in the X and Y directions will be described. An X-axis capacitance sensor 223 and a Y-axis capacitance sensor 222 fixed to at least one of the holders 201'and 202' by a mechanism (not shown) are provided, and the tip of the Z stage 214' The opposing surface on the Z stage is precision ground. The X-axis capacitance sensor 223 and the Y-axis capacitance sensor 222 are arranged on the X-axis and the Y-axis, respectively, on the XY plane including the visual field center position 212 (tip position of the probe 102) of the objective lens. ing. With this arrangement, the stage displacement (displacement of the tip position of the probe 102) at the center of the field of view 212 of the objective lens can be measured, and the yawing included in the operation of the X and Y stages on which the Z stage 214'is placed. Due to errors such as pitching, there is almost no Abbe error.

By performing feedback control using the output of these displacement meters, it is possible to accurately control the position of the tip of the probe to perform scanning and measure the shape and dimensions with high-precision SPM. By using a more accurate displacement sensor with a resolution of around 10 picometers, such as a laser interference displacement meter as described later, instead of the capacitance sensor, SPM measurement with even higher accuracy and resolution can be performed. realizable.

At this time, in order to realize high-speed measurement in a wide range and high-precision measurement in a narrow range, the following configuration is used. That is, instead of the piezoelectric element 205 described with reference to FIGS. 2A to 2C, a fine coarse movement drive mechanism consisting of 205a, 205b, 205c of FIGS. 4A and 4B is used. The same applies to the piezoelectric elements 206, 209, 210 if they are read as 205, so only 205 will be described (however, for the piezoelectric elements 209, 210, the Y stage 203 is read as the X stage 207, and the Y displacement sensor 222 is read as the X displacement sensor 223. Read as.). 205a is a piezoelectric element that constitutes a coarse motion mechanism. This is joined to the displacement expansion mechanism 205b. The 205b constitutes a lever, one end of which is fixed to the holder 201, the middle is pushed by the piezoelectric element 205a, and the other end is joined to the fine movement piezoelectric element 205c.

The displacement of the piezoelectric element 205a is expanded by the amount corresponding to the ratio of the distance between the fixed portion of the displacement expanding mechanism 205b and the joint between 205c and the distance between the fixed portion and the joint between 205a (lever enlargement ratio). It is transmitted to the element 205c. For example, if the lever magnification is 5 times and the elongation of the piezoelectric element 205a is 20 micrometers, one end of the piezoelectric element 205c will be displaced by 100 micrometers. Further, the piezoelectric element 205c transmits the displacement to the Y stage 204, and the Y stage is displaced. At this time, the displacement expansion system consisting of 206a, b, and c also transmits the displacement to the Y stage 204 in the same manner. As described above, if the displacement resolution of the piezoelectric elements 205a and 206a is 1 nanometer, the displacement noise transmitted to the stage 204 is also magnified 5 times to 5 nanometers.

Here, the piezoelectric elements 204a to d (see FIG. 2A), 205c, and 206c (see FIG. 4A) are piezoelectric elements for fine movement, and for example, those that extend by about 1 micrometer when a voltage of 100 V is applied are used. Then, the displacement of the fine movement piezoelectric element is added, and the Y stage is displaced. If the voltage noise is 5 mV, the noise of the piezoelectric element for fine movement is 50 picometers. At this time, if the response of the fine movement piezoelectric elements 204a to d or 205c, 207c is faster than that of the coarse movement piezoelectric elements 206, 206 or 205a, 207a, the Y stage 203 or 203'is the displacement sensor 222 and the Y stage 203 or 203'. By detecting the position in the Y direction and applying feedback, it is possible to control the Y stage 203 or 203'with a resolution of the same order as the resolution of the displacement sensor. In this feedback control, as will be described later with reference to FIG. 15, the stage displacement data detected by the stage displacement detection unit 128 is processed by the probe scanning control unit 112, and the piezoelectric element is driven via the probe drive control unit 110. To realize. The probe drive control unit 110 is equipped with a piezoelectric element driver 500 for coarse movement and a piezoelectric element driver 510 for fine movement for each axis. However, if the responsiveness on the coarse movement side is not sufficiently small, the noise cannot be canceled by the piezoelectric element for fine movement. That is, it is considered that the piezoelectric element is electrically a load having a capacitance C, and this is driven by a driver amplifier having an output resistance R. Then, when the capacitance of the driver amplifier is large and the output resistance R can be reduced, the time constant RC becomes shorter and the amplifier's input command position can be responded quickly, as shown in Fig. 5A. The noise is so large that it cannot be sufficiently canceled by the piezoelectric element for fine movement. Therefore, if the output resistance is increased as shown in Fig. 5B, the noise becomes smaller, but the time constant RC becomes larger instead, and there is a problem that the response becomes very slow when positioning a long distance that cannot be covered by the piezoelectric element for fine movement. It was.

Therefore, as shown in FIGS. 6A to 6F, the configuration of the roughing piezoelectric element driver 500 is shown. Therefore, the response switching mechanism is used to speed up the response of the coarsening piezoelectric element during long-distance positioning, and to increase the response during static operation. Slow down and cover the positioning with the piezoelectric element for fine movement. In FIG. 6A, the roughening piezoelectric element is represented by a capacitor 550 having a capacitance C. The driver amplifier 501 is output via the output resistor R (502) and is connected to the roughing piezoelectric element 550. A high-speed drive output resistor Rs (505) is connected to the output resistor 502 in parallel via a switch 506r. In the case of Rs &lt;&lt; R, when the responsiveness switching signal is turned ON, the output resistance becomes RsR / (Rs + R) Rs, and the response is performed at high speed with the time constant RsC. When the switch 506r is off, it responds slowly with responsive RC in a mode with less noise. For long-distance positioning, turn on the switch 506r, approach the target, and then turn it off to achieve both high-speed positioning operation and stability at rest.

In the example shown in Fig. 6A, the switching was explicitly performed, but it can also be performed automatically. As shown in Fig. 6B, insert the 506, which is connected in parallel with the diodes reversed, instead of the responsive switch 506r. Then, assuming that the forward voltage drop of the diode is Vd, when the current flowing through the output resistor 502 exceeds Vd / R, the diode is turned on, and the current flows through the high-speed response resistor Rs to speed up the response. When the voltage across the piezoelectric element 550 approaches the target value, the current decreases, so the switch 506 is turned off again, the response becomes low, and noise decreases instead. Alternatively, the same applies as shown in FIG. 6C. The resistor 502'of R-Rs is connected in series with the resistor Rs (505) to make an output resistor, and the above-mentioned switch 506 made of a diode is connected in parallel with the resistor 52'0. At this time, when the current flowing through 502'exceeds Vd / (R-Rs), the diode is turned on, and the current flows through the high-speed response resistor Rs to speed up the response.

As shown in FIG. 6D, the piezoelectric element driver 510 for fine movement is connected from the driver amplifier 511 to the piezoelectric element 551 having a capacitance Cp via the output resistor Rp (512). Rp should be determined so that the time constant RpCp can achieve the required responsiveness.

Figures 6E and 6F show the state of change in displacement with respect to the step-like position input at this time. As shown in Fig. 6E, the coarse displacement X responds at high speed with the time constant RsC at first, and when the difference between the command position and the displacement X becomes small, the response time constant drops to RC by the switch 506 or 506r and slowly. However, the position X changes when the displacement noise is small. As shown in Fig. 6F, the tremor displacement Xp is displaced at high speed with the time constant RpCp. However, since the movable range is small, the command position is displaced only within the movable range even if it exceeds that range, but the displacement noise is small.

Furthermore, as shown in Fig. 7, if a capacitor is inserted in series between the piezoelectric element for fine movement and the driver circuit, the voltage is divided by the ratio of the inverse of the capacitance of the piezoelectric element and the capacitor, and noise, that is, fine movement piezoelectric. The displacement resolution of the element is improved. In this figure, the driver amplifier 511 drives the piezoelectric element 551 through the output resistor 512 (resistance value Rp) and further through the capacitor 560 having a capacitance of 1 / α of the capacitance Cp of the piezoelectric element 551. As a result, the voltage across the capacitors 560 and 551 is divided into α to 1, and the noise is reduced to 1 / (1 + α). Here, the resistors αRb and Rb that are in parallel with the capacitor are high resistances on the order of MΩ such that the time constant RbCp is sufficiently long, which prevents the voltage division ratio from gradually shifting due to leakage of the capacitor charge. It is a working thing. For example, if α is 4, the displacement resolution in the above-mentioned example of the piezoelectric element for fine movement is 10 picometers, which is one fifth of 50 picometers. As a compensation, the movable range of the piezoelectric element for fine movement is 0.2 micrometer, which is one-fifth of 1 micrometer.

Next, the operation of the probe scanning control unit 112 of the present invention will be described with reference to FIGS. 8A to 8C. In Fig. 8A, the coarse motion target position 802 is given by open control without feeding back the measurement result of the displacement meter. After changing the coarse motion target position 802, the responsiveness of the responsiveness switching amplifier 806 is maintained at high speed by the responsiveness switching signal 801 until the response becomes stable, and then the responsiveness is switched to low speed (low noise). Of course, a configuration may be used in which automatic switching may be performed without inputting the responsiveness switching signal 801 from the outside. The position of the stage is detected by the displacement meter 811 and noise is reduced by applying an appropriate filter 812, and then compared with the target position 803 813, this error is transmitted to the amplifier 807 via the controller 805, and the fine piezoelectricity Feedback control of element 809. Here, what is called a controller is to apply an appropriate filter to the input, apply a gain, and output it, which is generally used in control theory. For example, a controller called a PID controller outputs the input itself, the integral of the input, and the derivative of the input, multiplied by different gains, and added together. With the configuration described above, both high-speed response over a wide range and high-precision positioning can be achieved at the same time.

As another configuration, as shown in Fig. 8B, the detection position 831 of the stage is filtered by 832 to reduce noise, and then compared with the target position 822, the position error is controlled by coarse and fine movements. A configuration is also conceivable in which the coarse-moving piezoelectric element 827 is driven via the responsiveness switching amplifier 826 via the devices 823 and 825, and the fine-moving piezoelectric element 829 is driven via the normal amplifier 828 for fine movement. Among them, the input to the fine movement amplifier 828 is applied to the control amount to the coarse piezoelectric element 827 via another controller 824. The operation of this controller increases the integral operation. Then, for example, if the fine-moving piezoelectric element 829 continues to be stretched on average for a long time, this is added to the control amount to the coarse-moving piezoelectric element 827, the coarse-moving piezoelectric element 827 is stretched, and the fine-moving piezoelectric element 829 is instead It will shrink. The above description holds true even when the expansion and contraction are reversed. This makes it possible to automatically adjust the displacement of the coarse-moving piezoelectric element so that the fine-moving piezoelectric element can be used in the center of the operating range on average. In the case of the embodiment in this figure, the responsive switching amplifier can fix the output value by the coarse-moving fixed signal, and when the SPM is used in the scanning range that can be covered by the fine-moving piezoelectric element, the coarse-moving piezoelectric element It is possible to explicitly stop the operation of and realize a lower noise position.

The state of displacement at this time will be described with reference to FIGS. 9A to 9C. When the command position changes significantly in a step-like manner, the position error is transmitted to the amplifier of the coarse-moving piezoelectric element and the amplifier of the fine-moving piezoelectric element via the respective controllers, and the coarse-moving displacement in FIG. 9A and the fine-moving displacement in FIG. 9B change. .. When the change in the command position is large, the coarse displacement in Fig. 9A responds at high speed and changes with the time constant RsC. On the other hand, the fine movement displacement Xp in Fig. 9B changes at high speed, but since this state is added to the input of the amplifier of the coarse-moving piezoelectric element via the controller, the displacement gradually shifts toward the coarse-movement displacement X, and the figure shows. Total displacement of 9C: As X + Xp approaches the command position, the tremor displacement returns to zero (intermediate position) again. The total displacement at this time is changed as shown in practice by adding the fine displacement displacement compared to the displacement shown by the dotted line when only the coarse-moving piezoelectric element is used. Therefore, the time constant changes at high speed according to the time constant RpCp of the fine displacement.

As another configuration, FIG. 8C shows that the operation of the coarse-moving piezoelectric element can be completely stopped at static state without controlling the coarse-movement fixed signal described in FIG. 8B, and the effect of further improving the displacement noise can be obtained. The configuration is also conceivable. The difference from Fig. 8B is that a dead zone is provided. The dead zone A842 is for preventing the coarse piezoelectric element 848 from being driven if the position error is within a certain range. For example, if the input is X, the output is Y, and the dead zone is ± W, Y = The output is represented by X + W (X &lt;-W), Y = 0 (-W &lt;X &lt;W), Y = XW (W &lt;X). By setting the dead zone to a range slightly smaller than the movable range of the fine-moving piezoelectric element 850, the coarse-moving piezoelectric element 848 can be automatically driven only when the fine-moving piezoelectric element 850 cannot be positioned. By setting the dead zone B846 in the same way, when the piezoelectric element 850 for fine movement continues to deviate from the center of the drive range, the integrated output of the controller 845 in front of the dead zone B846 increases, and the dead zone B Since the output of is turned on, the coarse-moving piezoelectric element can be driven only when the fine-movement piezoelectric element continues to be out of the center of the driving range.

The state of displacement at this time will be described with reference to FIGS. 10A to 10C. As shown in Fig. 10A, when the step-like change of the command position is small and smaller than the dead zone W set within the fine movement range, or when the coarse movement fixed signal is explicitly turned on, the coarse movement displacement X does not change. At this time, the position error is transmitted to the amplifier for the fine movement piezoelectric element in which the dead bands A and B are not in the signal path, and the fine movement displacement Xp responds with the time constant of RpCp as shown in Fig. 10B, as shown in Fig. 10C. It follows the command position at high speed as the total displacement.

Next, using FIG. 11A, as another configuration, a configuration in which the coarse-moving piezoelectric element 550 and the fine-moving piezoelectric element 551 are driven by only one amplifier is shown. The driver amplifier 501 drives a fine movement piezoelectric element (551) having a capacitance Cp via an output resistor Rp (512). The driver amplifier 501 is also connected to the roughing piezoelectric element 550 via the output resistor R (502). A high-speed drive output resistor Rs (505) is connected to the output resistor 502 in parallel via a switch 506r. In the case of Rs &lt;&lt; R, when the responsiveness switching signal is turned ON, the output resistance becomes RsR / (Rs + R) Rs, and the response is performed at high speed with the time constant RsC. When the switch 506r is off, it responds slowly with responsive RC in a mode with less noise. For long-distance positioning, turn on the switch 506r, approach the target, and then turn it off to achieve both high-speed positioning operation and stability at rest.

In the example shown in Fig. 11A, the switching was explicitly performed, but it can also be performed automatically. As shown in Fig. 11B, the diode is inserted in parallel instead of the responsive switch 506r in the opposite direction. Then, assuming that the forward voltage drop of the diode is Vd, when the current flowing through the output resistor 502 exceeds Vd / R, the diode is turned on, and the current flows through the high-speed response resistor Rs to speed up the response. When the voltage across the piezoelectric element 550 approaches the target value, the current decreases, so the switch 506 is turned off again, the response becomes low, and noise decreases instead. Due to the difference in responsiveness between the coarse movement amplifier and the fine movement amplifier, the fine movement piezoelectric element automatically responds to the positioning of the small stroke. For large stroke positioning, the coarse-moving piezoelectric element and the fine-moving piezoelectric element respond at the same time, and when the position approaches the target position, the response of the coarse-moving amplifier slows down, and the rest responds to the fine-moving piezoelectric element.

Next, using FIGS. 12A and 12B, another configuration that achieves both high-speed wide-range scanning of the probe 102 described in FIGS. 4A and 4 and precise scanning in a narrow range will be described. Since the configurations of FIGS. 12A and 12A and B are almost the same as the configurations described in FIGS. 2A to 2C, only the differences will be described. The Y-axis piezoelectric elements 205 and 206 are two types of coarse-moving piezoelectric elements 205a'and 206a' and fine-moving piezoelectric elements 205c' and 206c', respectively. In the configurations described in FIGS. 4A and 4B, the coarse motion was realized by increasing the displacement of the piezoelectric element, but in FIGS. 12A and 12B, it is realized by increasing the length of the piezoelectric element. The same applies to the X-axis, and the X-axis piezoelectric elements 209 and 210 are two types of coarse-moving piezoelectric elements 209a'and 210a' and fine-moving piezoelectric elements 209c' and 210c', respectively. The same applies to the Z-axis. As for the Z-axis piezoelectric elements 216 and 217, two types of coarse-moving piezoelectric elements 216a and 217a and fine-moving piezoelectric elements 216c and 217c are respectively transferred. By driving this with a circuit similar to that described with reference to FIGS. 4A and 4B, both high-speed scanning over a wide range and high-precision scanning over a narrow range can be achieved at the same time.

The sample stage 104 is a normal XYZ stage in this embodiment, but it goes without saying that the accuracy can be further improved by using the stage described in FIG. 4B.

Roughly, FIG. 13 will be used to describe yet another configuration that achieves both high-speed, wide-range scanning of the probe 102 described with reference to FIGS. 4A and 4B and precise scanning of a narrow range. A Y-axis coarse-moving piezoelectric element 205a'' is fixed to the holder 201''', which drives the Y stage 203''' via a displacement expansion mechanism 205b'' based on the principle of leverage. .. The Y stage 203 is supported with respect to the holder 201 so as to move smoothly only in the Y direction via the elastically deformed portions 204a', b', c', d'. The X coarse-moving piezoelectric element 210a'' is fixed to the Y stage 203''', which drives the X stage 207'' via the displacement expansion mechanism 210b'' based on the principle of leverage. The X stage 204'is supported with respect to the Y stage 203''' so as to move smoothly only in the X direction via the elastic deformation portions 208a', b', c', d'. Further, inside the X stage 207'', the fine movement stage 240'is supported so as to move smoothly only in the XY direction via elastic deformation portions 230 a', b', c', d'. Since the elastically deformed portions 230 a', b', c', d'are L-shaped, they are elastically deformed smoothly in both the X and Y directions.

The fine movement stage 240'is driven in the X and Y directions by the fine movement X piezoelectric element 210c'' and the fine movement Y piezoelectric element 205c'', respectively. The fine movement piezoelectric elements 210c'' and 205c'' are joined to the X stage 207'' and the fine movement stage 240'by an elastic hinge so that the force is transmitted only in the expansion and contraction direction of the piezoelectric element. The displacement of the probe in the 102XY direction is measured by the X displacement meter 223 and the Y displacement meter 222. By driving this with a circuit similar to that described with reference to FIGS. 4A and 4B, both high-speed scanning over a wide range and high-precision scanning over a narrow range can be achieved at the same time.

Roughly, FIG. 14 will be used to describe yet another configuration that achieves both high-speed wide-range scanning of the probe 102 described in FIGS. 4A and 4B and precise scanning in a narrow range. There is a coarse movement stage 241 in the holder 201'''', and the coarse movement stage 241 is supported so as to move smoothly only in the XY directions via the elastic deformation parts 231a, b, c, and d. The coarse movement stage 241 is driven in the X direction and the Y direction, respectively, by the X coarse movement actuator 210a''' and the Y coarse movement actuator 205a''', which are arranged externally. This actuator may be a large piezoelectric element, a voice coil motor, a rotary motor such as a servo motor or a step motor, and a rotary drive conversion mechanism such as a ball screw.

Further, inside the coarse movement stage 241, the fine movement stage 240'' is supported so as to move smoothly only in the XY directions via elastic deformation portions 230 a', b', c', d'. Since the elastically deformed portions 230 a', b', c', d'are L-shaped, they are elastically deformed smoothly in both the X and Y directions. The fine movement stage 240'' is driven in the X and Y directions by the fine movement X piezoelectric element 210c''' and the fine movement Y piezoelectric element 205c''', respectively. The fine movement piezoelectric elements 210c''' and 205c''' are joined to the coarse movement stage 241 and the fine movement stage 240'' by an elastic hinge so as to transmit a force only in the expansion and contraction direction of the piezoelectric element.

The displacement of the probe in the 102XY direction is measured by the X displacement meter 223 and the Y displacement meter 222. By driving this with a circuit similar to that described with reference to FIGS. 4A and 4B, both high-speed scanning over a wide range and high-precision scanning over a narrow range can be achieved at the same time. However, it is necessary to perform appropriate responsiveness switching or coarse motion fixation depending on the type of actuator 210a''', 205a'''. For example, a brake shoe (not shown) is pressed against a rod that pushes the coarse movement stage 241 using an electromagnetic brake to fix it by frictional force. Alternatively, the responsiveness can be reduced by attaching a fluid damper (not shown) to the rod that pushes the coarse movement stage 241 and reducing the flow path of the orifice of the damper. Alternatively, if an external coil is inserted in series between the voice coil motor coil and the driver amplifier and both ends of the external coil are short-circuited with a relay, a high-speed response is obtained, and if the relay is opened, the coil inductance increases. In proportion to the ratio, the response time constant of the current flowing through the coil, that is, the generated force of the voice coil motor, becomes longer, and as a result, the displacement noise is reduced.

Subsequently, the operation of the SPM according to the present invention will be described with reference to FIGS. 1, 15A and 15B. FIG. 15A is a diagram showing a part of the semiconductor manufacturing process using the SPM according to the present invention, and also describes the function of the in-line SPM. Although the following description assumes a semiconductor manufacturing process, the same applies to, for example, a hard disk manufacturing process other than semiconductors. In this case, the wafer may be read as a recording medium, a wafer, a rover obtained by cutting the wafer into strips, a head obtained by cutting the rover into hard disk head units, or the like.

Wafers processed in the order of manufacturing equipment A1501 and manufacturing equipment B1502 are divided into those that move to the processing of manufacturing equipment C1504 in units of one lot, and then move to the processing of manufacturing equipment C1504 after measuring with SPM1503. This ratio is instructed by the operator to the host computer in advance in consideration of the throughput of SPM1503 (the number of wafers processed per unit time). All manufacturing equipment 1501, 1502 and 1504 and SPM1503 are connected to the host computer 1505 of the semiconductor manufacturing line by a data network, and the host computer 1505 manages the history and processes of all wafers being manufactured. Further, the wafers are transported between the devices by a transfer device (not shown). For example, the manufacturing apparatus A1501 is a dry etching apparatus, the manufacturing apparatus B1502 is a resist stripping apparatus, and the manufacturing apparatus C1504 is a film forming apparatus.

The wafers that have been processed by the manufacturing apparatus B1502 are conveyed to the SPM1503 at a predetermined ratio based on the wafer process control information managed by the host computer 1505. The SPM1503 inquires the host computer 1505 about the process control information of the conveyed wafer, obtains the coordinate information of the measurement points on the wafer, and then performs the measurement. After the measurement is completed, the SPM1503 outputs the measurement results of each measurement point on the wafer to the host computer 1505, and the wafer is transferred to the manufacturing apparatus C1504 by the transfer device.

The host computer 1505 analyzes the measurement results obtained from the SPM1503, and changes (optimizes) the processing conditions of each manufacturing device 1501, 1502 and 1504 as necessary. For example, the SPM1503 measures the etching steps at a plurality of positions on the wafer, and the etching conditions of the manufacturing apparatus A1501 (dry etching apparatus) are changed based on the variation. Alternatively, as a result of analyzing the measurement result of SPM1503, the wafer may be returned to the manufacturing apparatus B1502 for reprocessing. The processing conditions in these cases are implemented under processing conditions different from the normal processing conditions in each manufacturing apparatus, and these conditions are appropriately determined and managed by the host computer 1505 based on the measurement results of the SPM1503. An operator may intervene in the above feedback work (depending on the situation, it may be feedforward that determines the processing conditions of the manufacturing process after SPM1503 based on the measurement result of SPM).

In the above flow, when determining (optimizing) the processing conditions of each manufacturing device based on the measurement results of SPM1503, the higher the measurement accuracy of SPM1503, the finer the processing conditions of each manufacturing device are set. can do. Further, in order to use the SPM1503 as an in-line device in the semiconductor manufacturing process, the wafer processed by the upstream manufacturing device of the SPM1503 is measured by the SPM1503 until it is processed by the next manufacturing device without passing through the SPM1503. Ideally, improving the throughput of inline SPM is an essential task.

FIG. 15B is a series of operation explanatory views of the SPM1503 according to the present invention. Hereinafter, the specific operation of the SPM1503 will be described together with FIG. Wafers that have been processed by the manufacturing devices 1501 and 1502 in the upstream process of SPM1503 are stored in a case in units of one lot and mounted on the wafer cassette of SPM1503 by the transfer device of the semiconductor manufacturing line (S1501). The SPM1503 reads the bar code on the wafer case and obtains the corresponding process information and inspection conditions from the host computer of the semiconductor manufacturing line (S1502). After that, the loader of SPM1503 takes out one wafer from the wafer cassette and mounts it on the sample stage 104 so that the orientation of the wafer orientation is constant (S1503).

Next, the wafer is aligned through the following steps (S1504). First, the wafer 103 is vacuum-sucked to the sample stage 104, and after reading the wafer number drawn on the surface of the wafer 103 with a detector (not shown), it moves directly under the probe drive mechanism 101 while being mounted on the sample stage 104. Will be done. The Z-axis direction position of the sample stage 104 at this time is performed at the bottom dead center. On the other hand, in the meantime, the observation optical system 105 has risen to the top dead point, and the objective lens 106 has been replaced by rotating a revolver (not shown). Will be replaced. The equifocal distances of the objective lens 106 and the alignment objective lens are the same. Subsequently, the observation optical system 105 is lowered, and the focus position is adjusted so that the focus position is on the back surface (upper surface) of the probe 102. This focusing operation is automatically performed by the image recognition of the optical image processing unit 108. Subsequently, the observation optical system 105 is further lowered by a fixed amount (for example, 1 mm), and the focal position of the observation optical system 105 is moved to a position lower than when the SPM image is taken.

The sample stage 104 moves in the XY direction to a position where the alignment mark position on the wafer 103 is within the field of view of the alignment objective lens (not shown), and then gradually rises in the Z direction to the observation optical system 105. The surface of the wafer 103 is aligned with the focal position, and the optical image processing unit 108 recognizes the alignment mark as an image. At this time, under the condition that the aperture ratio of the alignment objective lens is low, the probe 102 may be observed at the same time in the optical image obtained by the observation optical system 105. Therefore, it is desirable that the image recognition of the alignment mark is performed at a position within the field of view of the observation optical system 105 so that the alignment mark and the probe 102 do not overlap. The alignment marks on the wafer 103 are image-recognized at at least two places to obtain a correlation between the pattern on the wafer 103 and the XY coordinate axes of the sample stage 104, and are stored in the overall control unit 114.

During the alignment operation of the wafer 103, the observation optical system 105 is lowered by a fixed amount and the focal position is moved to a lower position than when the SPM image is taken, so that the surface of the wafer 103 and the tip of the probe 102 come into contact with each other. Never. After the alignment operation is completed, the observation optical system 105 rises to the top dead center again, and the revolver (not shown) is rotated to be replaced with the objective lens 106 having a high magnification (for example, 100 times). Then, the observation optical system 105 is lowered, and the focus position is adjusted so that the focal position is the back surface (upper surface) of the probe 102. This focusing operation is automatically performed by the image recognition of the optical image processing unit 108. Further, the observation optical system 105 is lowered by a fixed amount (for example, 1 mm), and the focal position is moved to a position lower than when the SPM image is taken. Although the operation of exchanging the objective lens at the time of aligning the wafer 103 has been described here, the observation optical system 105 may be provided with a zoom function for an optical image so that the observation magnification can be changed without exchanging the objective lens. Absent.

Based on the inspection information (coordinate information) obtained from the host computer, the overall control unit moves the sample stage 104 in the XY direction to a position where the first measurement point is within the field of view of the observation optical system 105 (S1505). The optical image processing unit recognizes the measurement points (or patterns around the measurement points) included in the field of view (display area on the TV monitor 107) of the observation optical system 105, and fine-tunes the XY axes of the sample stage 104. This allows the measurement points to be accurately positioned. The magnification of the objective lens 106 is 100 times, and the observation optical system 105 is lowered by a fixed amount (for example, 1 mm), the focal position is moved to a lower position than when the SPM image is taken, and the surface of the wafer 103 is observed. It is possible to observe the surface of the wafer 103 with high resolution without being affected by the presence of the probe 102 arranged in the field of view of the objective lens 106. For example, if the aperture ratio of the objective lens 106 is 0.7, the pattern on the wafer 103 can be clearly observed under the condition that the resolution is 1 micrometer or less. It is also possible to observe the pattern on the wafer 103 directly below the probe 102. This utilizes an optical phenomenon obtained under the condition that the aperture ratio of the objective lens 106 is large and the probe 102 occupies only a part of the field of view of the objective lens 106.

The position of the measurement point may be determined by the operator observing the TV monitor 107 and directly designating the coordinates from the overall control unit. After that, the observation optical system 105 is raised by a fixed amount, and the focus position is adjusted so that the focus position is on the back surface (upper surface) of the probe 102.

According to the SPM according to the present invention, it is not necessary to move the sample stage 104 from the determination of the measurement point in the field of view of the observation optical system 105 to the end of the subsequent measurement operation. In the conventional SPM, since the probe drive mechanism exists directly above the probe, the visual field position of the observation optical system and the SPM image measurement position are different. The operating time of the stage was required for positioning. Alternatively, even if the sample stage 104 has a function of observing the measurement point and the probe without moving the sample stage 104, the aperture ratio of the observation optical system is increased because the probe drive mechanism exists directly above the probe. It was not possible to observe the pattern on the wafer surface with sufficient resolution. According to the SPM according to the present invention, since the probe drive mechanism 101 is provided with a through hole 211, the sample stage 104 is not operated by using an objective lens having a high aperture ratio directly above the probe 102. It is possible to observe the measurement point and the probe.

Next, the operation of bringing the tip of the probe 102 into contact with the surface of the wafer 103 (S1506) will be described. The probe drive mechanism 101 is a three-dimensional (X, Y, Z) probe scanning mechanism having a structure in which a stage having an elastic deformation portion is driven by a piezoelectric element, and is held by a probe holder 115 at the bottom thereof. The probe 102 is attached. The probe drive mechanism 101 is provided with a through hole for inserting the objective lens 106 in a non-contact manner, and the sample stage 104 can be moved by adjusting the focus axis (not shown) of the observation optical system 105. It is possible to observe the surfaces of the probe 102 and the wafer 103 without any problem. The movable area of the probe drive mechanism 101 is 20 micrometers in the X-axis direction, 20 micrometers in the Y-axis direction, and 10 micrometers in the Z-axis direction, and the detailed structure thereof is as shown in FIGS. 2A to 2C.

The contact between the tip of the probe 102 and the surface of the wafer 103 will be described with reference to FIGS. 16A and 16B. While monitoring the detection signal in the probe deflection detector 109, (1) raise the height of the Z-axis of the probe drive mechanism 101 to top dead center, and (2) raise the Z-axis of the sample stage 104 by 10 micrometers. (3) This is achieved by repeating the process of lowering the height of the Z-axis of the probe drive mechanism 101 to the bottom dead center. That is, when the tip of the probe 102 comes into contact with the surface of the wafer 103 in the process of (3) described above, the detection signal of the probe deflection detection unit 109 changes. The probe scanning control unit 112 detects this change and detects the contact between the two, and the detailed operation principle is as described in FIG.

In FIG. 16B, the probe 102 is in contact with the sample 103 at the third time of the probe descent in (3). After that, after raising the probe 102 to the top dead center, the amount calculated so that the sample surface height is within the target height 912 set near the center of the Z-direction movable range 911 of the probe 102. Only raise the Z-axis of sample stage 104. After that, the probe 102 is lowered again until the contact between the tip of the probe and the surface of the wafer 103 is detected (S1507). In the above method, when the probe and the sample come into contact with each other, only the probe 102, which can be controlled with higher accuracy, is driven, so that the probe 102 is driven by the sample 103 when the Z-axis on the sample side is driven. It has the effect of preventing damage to the tip of the probe, which may occur due to excessive force being applied to the probe.

A method of approaching another probe 102 and the sample 103 will be described with reference to FIGS. 17A and 17B. Here, the sample 103 is raised by the sample Z stage with the probe 102 lowered to the bottom dead center. As soon as the contact between the sample 103 and the probe 102 is detected, the probe 103 is quickly retracted to the top dead center. As a result, the state in which the surface of the sample 103 vibrating due to driving and the probe 102 are in contact with each other is minimized, and damage to the tip of the probe 102 is prevented. The surface of the sample 103 is set near the center of the movable range 911 of the probe 102, and is stopped after continuing to rise by the amount calculated so that the height of the sample surface comes within the target height 912. Then, the probe 102 is lowered again until the contact between the tip of the probe and the surface of the wafer 103 is detected.

After detecting the contact between the tip of the probe 102 and the surface of the wafer 103, the probe drive mechanism 101 is driven to scan the probe 102 and collect an SPM image. For example, a region of 1 micrometer square on the wafer 103 is divided into 256 in the X direction and 10 in the Y direction, the probe 102 is raised by, for example, 1 micrometer, and then the contact position is set in the X direction (Y direction). The contact detection between the probe 102 and the surface of the wafer 103 is repeated by sequentially moving the probe 102. However, this contact detection is performed by operating only the probe 102 within the movable range of the probe drive mechanism 101 without moving the sample stage 104. The operation of the probe drive mechanism 101 is controlled by the probe scanning control unit 112 via the probe drive control unit 110. Capacitance sensors are attached to each moving axis (X, Y, Z stage) of the probe drive mechanism 101 as shown in FIGS. 11A and 11B, and the displacement of each capacitance sensor is the stage displacement detector. It is detected by 128 and stored in the SPM image generation unit 114 via the probe scanning control unit 112. The SPM image generator 114 generates an XY plane distribution image of the displacement of the probe 102 (Z stage displacement of the probe drive mechanism 101) measured in a state where the probe 102 is in contact with each contact point on the wafer 103. To do. The piezoelectric element used in the probe drive mechanism 101 can operate at a response speed of 2 to 3 kHz, the above measurement operation is completed in a few seconds, and the obtained SPM image (data) is the whole. It is saved in the control unit.

The measurement coordinates and the number of measurement points on the wafer 103 are predetermined. If other measurement points remain on the wafer 103, the observation optical system 105 and the sample stage 104 are lowered by the same amount, and the XY of the stage 104 is lowered. Move the coordinates to the coordinates of the next measurement point and move to the measurement operation again (S1508). If there are no other measurement points on the wafer 103, the observation optical system 105 and the sample stage 104 are lowered by the same amount, and the wafer 103 is unloaded from the sample stage 104 (S1509). If the next wafer to be measured exists in the wafer case, it is loaded on the sample stage 104 and the measurement is repeated (S1510). When the measurement of all the wafers in the wafer case is completed, the overall control unit 114 is notified. The stored data is output to the host computer (S1511), and a transfer device (not shown) transfers the wafer case to the next processing device (S1512).

In the conventional SPM, every time the measurement position is changed on the same wafer, it is necessary to recognize the measurement position by the observation optical system that performs measurement positioning and move the sample stage to the probe scanning mechanism (position of the probe 102). There was. When the distance between the observation optical system position for measurement positioning and the probe scanning mechanism (position of the probe 102) is 150 mm, it takes 2 to 3 seconds for the sample stage to operate, but one sheet. When 10 measurement points exist on the wafer, the total operating time of the sample stage for the measurement positioning operation is 20 to 30 seconds, which is a factor that greatly reduces the measurement throughput of SPM. According to the operation of the SPM of the present invention, the surface of the probe 102 and the wafer 103 can be observed without moving the sample stage 104 by adjusting the focus of the observation optical system 105, so that the time required for the measurement positioning operation can be omitted. can do. As a result, the time required for a series of device operations (loading the wafer, measuring the etching step at 9 measurement positions, and then unloading the wafer) for each wafer to be measured is 2 minutes or less (30 WPH). Inline SPM with improved throughput can be realized.

Next, using FIGS. 18A and 18B, the operation of the present invention is performed when the measurement position is not specified by the coordinates of the sample XY stage 410 or the observation result by the optical microscope 105, but is precisely specified by the SPM measurement result. explain. Reference numeral 901 is a scannable range of the probe 102 by the coarse-moving piezoelectric element. Among them, the range 910 that seems to include the pattern to be measured is first scanned by the probe 102 to obtain the measurement result 910. In this, the pattern to be measured is found, and the area including only this pattern is precisely measured by the probe 102. At this time, if this area is smaller than the measurable size by the fine-moving piezoelectric element (fine-moving piezoelectric element scannable area) 902, fix the coarse-moving piezoelectric element and perform high-precision measurement by scanning only the fine-moving piezoelectric element. Can be done. Regarding the Z axis, if two types of actuators, coarse motion and fine motion, are provided, the position of the coarse motion piezoelectric element is adjusted so that the height of the area to be measured falls within the height control width 912 by the fine motion piezoelectric element. By measuring after that, it is possible to measure the height with high accuracy.

This measurement sequence will be described with reference to FIG. First, move the sample stage and bring the probe 102 to the position on the sample 103 that you want to measure (S1901). Next, if it is necessary to explicitly switch the responsiveness of the Z coarse-moving amplifier, switch it at high speed (S1902). Next, the sample approach processing described with reference to FIGS. 16 and 17 is performed (S1903). Next, if it is necessary to explicitly switch the responsiveness of the XY coarse-moving amplifier, switch it at high speed (S1904). Next, the coarse motion scanning of the probe 102 is performed to perform measurement (S1905) (the above control is performed by the probe scanning control unit 112 of FIG. 15). The position of the pattern to be measured is automatically obtained by image processing in the overall control unit 114 from the measured images obtained by this (in the SPM image generation unit 113), or the user specifies on the screen (S1906). .. After moving the XY axis to the center of the area under test (S1907), if the responsiveness of the XY coarse-moving amplifier needs to be explicitly switched, switch it to low speed (S1908). Alternatively, coarse motion fixing is performed (S1909). If the Z-axis also has a coarse / fine movement mechanism, perform the same processing. After that, scanning is performed by the finely moving piezoelectric element of the probe 102 to realize high-precision measurement at a high magnification (S1910). Here, if the measurement is to be completed (S1911), if it is necessary to explicitly switch the responsiveness of the Z coarse-moving amplifier, switch this at high speed (S1912), and then save the sample 103 (S1913). ). If you want to zoom further or measure the measurement position again within the field of view by the same fine movement scanning, perform the fine movement scanning again.

If you want to measure in a different field of view (S1920), re-do the position from where you want it by image processing or user specification. If there is no pattern to be measured within the same coarse scan range, if it is necessary to explicitly switch the responsiveness of the XY coarse amplifier again, switch this at high speed (S1921), and then rough scan. Change the range and start scanning again (S1922).

By the sequence described above, it is possible to realize high-speed and high-precision measurement by switching between high-speed scanning over a wide range obtained by the present invention and high-precision scanning over a narrow range.

As the displacement meter used in the present invention, an example of a laser interference displacement meter different from the capacitance sensor will be described with reference to FIGS. 20 and 21A to 21C. In the case of the capacitance sensor, the distance between the target and the flat metal electrode was converted into the change in capacitance and detected, but in the case of the laser interference displacement meter, the distance from the flat mirror is the interference fringe. It is detected by converting it to the phase of.

As shown in FIG. 20, the optical interference displacement sensor of this embodiment includes a light source unit (not shown), a sensor unit 100, and a displacement output unit 70. In the light source unit, for example, linearly polarized laser light having a wavelength of 632.8 nm from a frequency-stabilized He-Ne laser is guided to the sensor unit 100 by the polarization plane preservation fiber 2 in a polarization direction of 45 °.

The sensor unit 100 includes an interferometer 600 and a displacement calculation processing unit 50. In the interferometer 600, the 45 ° polarization emitted from the polarization plane preservation fiber 2 is converted to parallel light 4 by the collimator 3, and further transmitted through a polarizing element 5 such as a Gran Thomson prism, and the transmitted light 6 is transmitted to the prism mirror 7 and the unpolarized beam splitter. It is reflected by 8 and incident on the reference mirror 9.

As shown in FIG. 21B, the reference mirror 9 has a configuration in which a diffraction grating 9b is formed on a synthetic quartz substrate 9a with a metal material such as aluminum. The polarized beam 6 in the 45 ° direction incident on the diffraction grating consists of two vector-decomposed orthogonal polarization components, and the S polarization component 25s parallel to the longitudinal direction of the diffraction grating is reflected by the diffraction grating and is orthogonal to the P polarization component. 25p passes through the diffraction grating. That is, this diffraction grating exhibits properties as a so-called diffraction polarizing element (Wire Grid Polarizer). In the case of this example, the pitch of the diffraction grating 9b was 144 nm, the line width was 65 nm, and the height was 165 nm.

The S-polarized beam 6r reflected by the reference mirror 9 is used as the reference light. The transmitted P-polarized beam 6m is used as measurement light. The P-polarized beam 6m passes through the 1/4 wave plate 10 and then becomes circularly polarized light, is reflected by the target mirror 12 placed on the measurement object 31, and once again passes through the 1/4 wave plate 10 and then becomes S-polarized light. It is reflected by the reference mirror 9, transmitted through the 1/4 wave plate 10 and then reflected by the target mirror 12 as circularly polarized light, transmitted through the 1/4 wave plate 10 and then converted to P-polarized light, and transmitted through the reference mirror 9. That is, the measurement light 6m makes two round trips in the optical path between the reference mirror 9 and the target mirror 12, and the movement amount 31d of the measurement object 31 is doubled and detected. The S-polarized beam 6r reflected by the reference mirror 9 and the transmitted P-polarized beam 6m are combined as an orthogonally polarized beam 14 and transmitted through the unpolarized beam splitter 8.

After passing through the opening 13 for removing stray light, the orthogonally polarized beam 14 is divided into four orthogonally polarized beam beams 17 by two pyramid-shaped quadrangular pyramid prisms 15a and 15b facing each other. The method of beam division is not limited to such a prism, and a diffractive optical element or the like can also be applied. The four orthogonally polarized beam beams 17 transmit polarization interference through the phase shift elements 18 and 19 so that the phase shifts of 0, π / 2, π, and 3π / 2 are given between the orthogonal polarization components. Then, four phase shift interference lights 20 are generated.

The phase shift element 18 is divided into two as shown in FIG. 21A, the lower half is composed of synthetic quartz 18d, and the upper half is composed of photonic crystal 18c. As shown in the enlarged view, the photonic crystal 18c forms a horizontal line-and-space diffraction grating with a pitch smaller than the wavelength of the incident light on the synthetic quartz substrate 18c1, and dielectrics having different refractive indexes are formed on the diffraction grating. It is composed of laminated thin films 18c2 and 18c3. The cross section of the thin film layer deposited on the diffraction grating maintains a triangular waveform uneven shape in the film thickness direction due to the unevenness of the diffraction grating. As the thin film material, Si, SiO2, TiO2, Ta2O5, and Nb2O5 can be applied. Such a multilayer thin film structure based on a diffraction grating becomes a photonic crystal whose crystal axis direction is the direction of the diffraction grating, exhibits double refraction characteristics due to diffraction and interference between the multilayer thin films, and polarizes or transmits incident light. -It is possible to control the reflection characteristics (see: Photonic Lattice Product Catalog Co., Ltd.). The diffraction grating pitch, depth, and film thickness of each thin film are controlled in consideration of the wavelength of the incident light and the desired characteristics. Another major feature is that polarizing elements and wavelength elements with different crystal axis directions can be formed in an array on a single substrate by using photolithography technology used in semiconductor element manufacturing and film attachment technology such as sputtering. is there. The photonic crystal 18c has a function as a 1/4 wave plate, and a thick arrow indicates the crystal axis direction. That is, as shown in FIG. 19, of the four orthogonally polarized beams 17, two orthogonally polarized beams that pass through the photonic crystal 18c have a phase difference of π / 2 between the two polarized components. On the other hand, the remaining two orthogonally polarized beams pass through the synthetic quartz 18d and there is no phase difference.

As shown in FIG. 21C, the phase shift element 19 is divided into two parts, the left half is composed of a photonic crystal 19a having a crystal axis direction of 45 °, and the right half is a photonic crystal 19a having a crystal axis direction of 45 ° in the opposite direction. It is composed of nick crystals 19b. Similar to the photonic crystal 18c, the photonic crystal 19a forms a line-and-space diffraction grating in the 45 ° direction with a pitch smaller than the wavelength of the incident light on the synthetic quartz substrate 19a1 as shown in the enlarged view. It is composed by laminating dielectric thin films 19a2 and 19a3 having different refractive indexes on top of each other. The structure of the photonic crystal 19b is similar. The photonic crystals 19a and 19b have a function as a polarizing element, and a thick arrow indicates the crystal axis direction thereof. That is, as shown in FIG. 20, of the four orthogonally polarized beam beams 17, the two polarized components constituting the two orthogonally polarized beams transmitted through the photonic crystal 19a and the two orthogonally polarized light transmitted through the photonic crystal 19b. Both polarizing components interfere with each other in a state where a relative π phase difference is given between the two polarizing components constituting the beam.

That is, polarization interference occurs in a state where a phase shift of 0, π / 2, π, or 3π / 2 is given between each orthogonal polarization component of the four orthogonally polarized beams 17 that have passed through the phase shift elements 18 and 19. Four phase shift interference lights 20 are generated. In order to avoid the influence of ambient light, the four phase shift interference lights 20 pass through the interference filter 21 having a transmission center wavelength at a wavelength of 632.8 nm, and then are received by the four photoelectric conversion elements such as photodiodes at 22 respectively. After being amplified by the amplifier 23, it is output as four phase shift interference signals 41a, 41b, 41c, and 41d.

The four phase shift interference signals 41a, 41b, 41c, and 41d are given by Eqs. (1) to (4), respectively. Ia = Im + Ir + 2 (Im Ir) 1 / 2cos (4πnD / λ) (Equation 1) Ib = Im + Ir + 2 (Im Ir) 1 / 2cos (4πnD / λ + π) = Im + Ir-2 (Im Ir) 1 / 2cos (4πnD / λ) (Equation 2) Ic = Im + Ir + 2 (Im Ir) 1 / 2cos (4πnD / λ + π / 2) = Im + Ir + 2 (Im Ir) 1 / 2sin (4πnD / λ) (Equation 3) Id = Im + Ir + 2 (Im Ir) 1 / 2cos (4πnD / λ + 3π / 2) = Im + Ir-2 (Im Ir) 1 / 2sin (4πnD / λ) (Equation 4) Here, Im is the detection intensity of the probe light, Ir is the detection intensity of the reference light, n is the refractive index of air, D is the movement amount 31d of the object to be measured 31, and λ is the wavelength of the laser light 4.

In the displacement calculation processing unit 50, the movement amount D of the measurement object 31 is calculated based on the equation (5), and is displayed on the displacement output unit 70 as the movement amount signal 61. D = (λ / 4πn) tan-1 {(Ic-Id) / (Ia-Ib)} (Equation 5) In this embodiment, a diffraction polarizing element (Wire Grid Polarizer) is used as the reference mirror 9, but as is clear from the above description, a photonic crystal 9c having a crystal axis direction in the horizontal direction as shown in FIG. 21C is used. It can also be used. Similarly, for the 1/4 wavelength plate 10, it is also possible to use a photonic crystal 10c having a crystal axis direction in the 45 ° direction. Further, in order to further simplify the interferometer 600, the phase shift element 19 is composed of only the photonic crystal 19a in FIG. 21C, and the phase shift interference signals 41a and 41c represented by the equations (1) and (3) are used. It is also possible to obtain the movement amount D of the measurement object 31 from these two interferometric signals.

As is clear from FIG. 20, the two beams of the measurement light 6m and the reference light 6r toward the target mirror 12 are emitted from the light source unit and incident on the sensor unit 100, and further reach the reference mirror 9 until the reference mirror 9 is reached. It passes through exactly the same optical path from the light to the light received by the four photoelectric conversion elements 22. That is, it has a configuration of a common optical path type interferometer. Therefore, even if a temperature distribution, a refractive index distribution, or mechanical vibration occurs in the optical path due to air fluctuations, these disturbances affect both beams equally. Therefore, when both beams interfere with each other, these disturbances occur. The effects are completely offset and the interfering light is unaffected by disturbances. Only the measurement light of 6 m exists in the optical path between the reference mirror 9 and the target mirror 12, but since the stroke of a scanning probe microscope or the like is at most several hundred microns, the reference mirror 9 and the target mirror 12 have a stroke of at most several hundred microns. The gap with and can be set to 1 mm or less, and the influence of disturbance in such a minute gap can be ignored. Further, the fluctuation of the intensity of the laser beam itself is the fluctuation of the probe light detection intensity Im and the reference light detection intensity Ir in the equations (1) to (4), but the subtraction in the equation (5) in the displacement calculation processing unit 50. It is offset by processing and division processing.

Further, in the optical interference displacement sensor of this embodiment, four orthogonally polarized beams are generated with a simple configuration, and four phase shift interference lights are generated and received in parallel spatially by phase shift elements arranged in an array. Since it has a configuration, there is an advantage that the number of optical components is significantly reduced and the displacement sensor is remarkably miniaturized as compared with the conventional phase shift interferometer. Specifically, the size of the interferometer 600 can be reduced to about 20 × 15 × 50 mm or less. In addition, since the four phase shift interference lights pass through the adjacent optical paths, even if disturbances such as temperature distribution, humidity distribution, pressure distribution, density distribution, and airflow change due to air fluctuations are superimposed between the optical paths, The impact can be minimized.

From the above, the optical interference displacement sensor of this embodiment can move the amount and position of the object to be measured from sub-nanometer to 10 picometers without controlling environmental factors such as temperature, humidity, pressure, density, and acoustic vibration with high accuracy. It is possible to measure stably with the following accuracy, and by using this to feedback control the SPM scanning mechanism, the position of the SPM probe tip can be stably controlled with an accuracy of 10 picometers or less from sub-nanometers. It is possible to realize a highly accurate SPM device.

1 ... Light source unit 2 ... Wave plate preservation fiber 3 ... Collimator 5 ... Polarizing element 7 ... Prism mirror 8 ... Unpolarized beam splitter 9 ... Reference mirror 10 ... 1 / 4 wave plate 12 12x 12y 12z ... Target mirror 14 17 81 218 ... Orthogonal polarization beam 13 ... Aperture 15a 15b ... Square pyramid prism 18 19 82 83 84 219 220 ... Phase shift element 9c 10c 18c 19a 19b 82c 82d 83c 83d 84a 84b ... Photonic crystal 20 85 221 ... Phase shift interference light 21 86 ... Interference filter 22 ... Photoelectric conversion element 23 88 223 ... Amplifier 31 .. Object to be measured 50, 51, 52 ... Displacement calculation processing unit 70 ... Displacement output unit 101 ... Probe drive mechanism 102 ... Probe 103 ... Wafer 104 ... Sample stage 105. .. Observation optical system 106 ... Objective lens 201, 202 ... Holder 203 ... Y stage 204a 204b 204c 204d ... Elastic deformation part 205 206 209 210 216 217 ... Piezoelectric element 205a 206a 209a 210a 216a 217a ... Coarse Dynamic Piezoelectric Element 205b 206b 209b 210b ... Displacement Enlargement Mechanism 205c 206c 209c 210c 216c 217c ... Fine Motion Piezoelectric Element 205a '210a ... Coarse Motion Actuator 207 ... X Stage 207 ... Sample Stage 211 ... Through hole 214 ... Z stage 240 ... Fine movement stage 230a 230b 230c 230d ... Elastic deformation part 231a 231b 231c 231d ... Elastic deformation guide 241 ... Coarse movement stage 220 221 ... Target 222 223 224 410 ... Displacement sensor 301 ... Laser diode 311 ... Receiver 402 ... Coarse XY stage 404 ... Air slider 403 ... Coarse Z stage 405 ... Surface plate 406 ... Elastic plate 410 ... Sample XY stage 500 ... Piezoelectric element driver for coarse movement 501. .. Driver amplifier 506 ... Switch 506 r ... Switch r 510 ... Piezoelectric element driver for fine movement 511 ... Driver amplifier 512 ... Output resistance.