JP2006300676A - Flatness anomaly detecting technique and exposing device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a flatness anomaly detecting technique with good throughput capable of detecting anomalies of flatness by examining the test substance during short time. <P>SOLUTION: Utilizing an automatic focus detecting system 6 of the exposure device 1, a positional information acquiring step acquiring positional information while moving an wafer W and a projected slit image relatively, an approximate flat surface acquiring step comparting an wafer W to shot domains to derive an approximate flat surface from positional information contained there, a flatness abnormality detecting step, where the approximate surface is compared with each positional information, the disjunction from approximate flat surface is also compared with a threshold, and the case the disjunction is larger than this threshold is discriminated as flatness abnormality, are all implemented and the shot domain is displayed on a displaying device 20c the abnormality is detected. When dust exists on the wafer W, the dust is detectable by a simple procedure prior to exposure process. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a flatness abnormality detection method and an exposure apparatus.

  In recent years, due to miniaturization of semiconductor device rules, an exposure apparatus is required to have a high NA. If the NA is increased, the resolution of the projected image is improved, but there is a problem that the depth of field becomes shallow. Therefore, in order to achieve an accurate focus, for example, a multipoint focal position detection system having a plurality of points on a photosensitive substrate such as a wafer as detection points is provided. According to this multipoint focal position detection system, slit images can be projected at a plurality of points on the photosensitive substrate, and position information in the optical axis direction of the projection optical system can be detected from the reflected light. For this reason, since not only the position of the photosensitive substrate surface in the optical axis direction but also its inclination can be detected, the area to be exposed on the photosensitive substrate can be more accurately focused on the image plane of the projection optical system. .

  In this case, in order to focus accurately, high-accuracy leveling (tilt correction) is a prerequisite. Even if an attempt is made to detect an inclination from a difference in height at close measurement points, the inclination cannot be detected with high accuracy. In order to perform leveling with high accuracy, as shown in Patent Document 1, the heights of measurement points at distant positions are continuously measured, the heights are averaged, and the inclination is detected based on this average. Leveling.

  Further, when the wafer is vacuum-sucked on the wafer stage, the flatness of the wafer is affected by the flatness of the wafer holder surface. On the other hand, since the wafer is scanned and stepped at a high speed, the wafer is supported by levitation by an air bearing or the like, and an unexpected inclination may occur due to the indefiniteness of the running surface. In order to measure the flatness in such a state, not only the inside of the shot area as described in Patent Document 2, but also the shot area from the position information in a wide range using detection points provided outside the shot area. The surface position information is obtained and the tilt is detected with high accuracy. In this way, there is one that measures the flatness of the shot area without being affected by the indefiniteness of the running surface.

According to such a method, it is possible to accurately measure the flatness of the wafer surface even on a wafer placed on a wafer stage that is levitated and supported by an air bearing.
JP-A-6-5495 JP 2004-138611 A

  However, the exposure result only affects the wafer tilt, indeterminacy due to wafer running, the thickness of the wafer itself, the polishing state, the flatness of the surface of the wafer holder on which the wafer is placed, etc. is not. Of these, the most important are foreign matter on the wafer and foreign matter between the wafer holder and the wafer, that is, dust. These need to be excluded because the exposure result is seriously affected by blocking exposure light or causing large deformation locally.

  In the case of measuring only the inclination and flatness as described in Patent Document 1 at a high speed, a fine foreign substance having a large sampling position interval cannot be found for that purpose. In addition, as shown in Patent Document 2, the step for each shot region and the inspection of a large number of measurement points inside and outside the shot region repeat the inspection itself when it is attempted to inspect the entire wafer in detail. There was a problem that the throughput deteriorated. Moreover, although the thing shown in patent document 2 can test | inspect the whole wafer with high precision, and can measure a precise flatness and can detect a foreign material, it took time for the whole test | inspection, and the human finally needed to discriminate | determine .

  Originally, the multipoint focal point position detection system is designed to measure the inclination by measuring at a distant position. There was also a problem of being low.

  In order to solve the above problems, an object of the present invention is to provide a flatness abnormality detection method and an exposure apparatus with good throughput that can inspect an object to be inspected in a short time and detect an abnormality in flatness.

Another object of the present invention is to provide a flatness abnormality detection method and an exposure apparatus that can detect a fine flatness abnormality without leaking and can manufacture a highly reliable product.
Another object of the present invention is to provide a flatness abnormality detection method and an exposure apparatus capable of easily confirming an abnormality in flatness from measured numerical values.

In order to solve the above-described problem, in the flatness abnormality detection method according to the present invention, a plurality of pieces of position information in the height direction (Z-axis direction) of the surface of the inspection object (W) are set on the surface of the inspection object. A flatness abnormality detection method for detecting an abnormality in flatness of a surface of the object to be inspected in a region (S22 to S28), wherein the position information is obtained while relatively moving the object to be inspected and the measurement region. Detecting and obtaining the position information at a plurality of measurement points (MP) on the surface of the object to be inspected, and dividing the surface of the object to be inspected into a plurality of inspection regions (SHn); An approximate plane acquisition step for obtaining an approximate plane (APn, APn ′) of the inspection area from the position information of the plurality of measurement points included in one inspection area of the plurality of inspection areas; Inspection area Each of the position information at the plurality of measurement points included in the area is compared with the approximate plane, and the deviation from the approximate plane is compared with a first threshold value (K ₂ ). The gist of the present invention is that it includes a flatness abnormality detection step of extracting the inspection region as a flatness abnormality when it is larger than the threshold value.

  In the present invention, since the position information is detected while relatively moving the object to be inspected and the measurement area, the position information can be acquired continuously at a relatively high speed. In addition, since an approximate plane is generated and abnormality in flatness is determined based on this, a more accurate determination can be made. In addition, since an inspection region having an abnormality in flatness is automatically extracted, there is an effect that troubles caused by dust or the like can be reduced extremely easily and quickly.

In the exposure apparatus according to the present invention, a plurality of measurement areas (S22 to S28) are formed by irradiating the surface of the object to be inspected (W) with a detection light beam (DL), and the surface of the object to be inspected in the measurement area A detecting device (6) for detecting position information in the vertical direction (Z-axis direction), a moving device (26) for relatively moving the object to be inspected and the detecting device, and controlling the moving device and the detecting device. An exposure apparatus (1) comprising a control device (20) for detecting an abnormality in the flatness of the surface of the object to be inspected, the control device during the relative movement between the object to be inspected and the measurement region Position information acquisition means for detecting the position information and obtaining the position information at a plurality of measurement points (MP) on the surface of the object to be inspected, and partitioning the surface of the object to be inspected into a plurality of inspection regions (SHn) And the plurality of inspection areas Approximate plane acquisition means for obtaining an approximate plane (APn, APn ′) of the inspection area from the position information of the plurality of measurement points included in one inspection area, and the plurality of the plurality of points included in the one inspection area. Each of the position information at the measurement point is compared with the approximate plane, and the deviation from the approximate plane is compared with the first threshold value (K ₂ ), and flat when the deviation is larger than the first threshold value. The gist of the invention is to have a flatness abnormality detecting means for extracting the inspection region as a degree of abnormality.

  In the present invention, since position information is detected while relatively moving an object to be inspected such as a wafer and a measurement region, position information can be acquired continuously at a relatively high speed, and inspection has an abnormality in flatness. Since the area is automatically extracted, trouble caused by dust can be reduced extremely easily and quickly. In particular, since relatively high-speed processing is possible, it can be performed at the start of device production or during production, and there is an effect that a highly reliable device can be manufactured with high productivity.

  In addition, in order to explain each invention clearly, it demonstrated corresponding to the code | symbol of drawing which shows one Example, However, It cannot be overemphasized that this invention is not what is limited to an Example.

  The present invention has an effect that it is possible to detect an abnormality in flatness by inspecting an object to be inspected in a short time.

  Embodiments in which the present invention is embodied in an exposure apparatus will be described below. In the exposure apparatus 1 of the present embodiment shown in FIG. 1, an image of an object such as a mask or a reticle (hereinafter, representatively described by a reticle R) is represented by an object such as a silicon wafer or a glass substrate (hereinafter, represented by a wafer W). This is a step-and-scan type reduced projection exposure apparatus (scanner) that projects and exposes on the projection. The object such as the wafer W corresponds to the inspection object of the present invention.

  The exposure apparatus 1 holds an illumination system 12 that emits illumination light IL, a reticle stage RST that holds a reticle R, a projection optical system PL that projects an image of the reticle R, and a wafer W that forms an image by the projection optical system PL. A wafer stage WST is provided. The detecting device and the moving device constituting the flatness abnormality detecting device of the present invention are arranged on this wafer stage WST. Moreover, the control system centering on the main controller 20 which controls these is provided.

  Although not shown in detail, the illumination system 12 includes a light source that emits a light beam, an illumination optical system including a collimator lens, an illuminance uniformizing optical system including an optical integrator, a relay lens system, a reticle blind, a condenser lens, and the like. It is composed of

In this embodiment, the light source includes, for example, an ArF excimer laser oscillator (wavelength 193 nm). The light source is not limited to this, but a KrF excimer laser oscillator (wavelength 248 nm), an F ₂ laser oscillator (wavelength 157 nm), a vacuum ultraviolet light source that outputs ultraviolet pulsed light in the vacuum ultraviolet region, G-line / I A line or the like may be used.

As the optical integrator, a fly-eye lens, a rod integrator (an internal reflection type integrator), a diffractive optical element, or the like can be used.
According to the illumination system 12, the illumination light IL for the exposure light EL generated by the light source passes through a shutter (not shown), and is then converted into a light flux having a substantially uniform illuminance distribution by the illuminance uniformizing optical system. The illumination light IL emitted from the illuminance uniformizing optical system reaches the reticle blind via the relay lens system. The light beam that has passed through the reticle blind passes through a relay lens system, a condenser lens, and a reticle R on which a circuit pattern or the like is drawn. Illumination area) Illuminates the IAR with a uniform illuminance distribution.

  On reticle stage RST, reticle R is fixed, for example, by vacuum suction. Reticle stage RST can be slightly driven in the Z direction parallel to the optical axis of illumination system 12 (corresponding to optical axis AX of projection optical system PL described later) by a reticle drive system (not shown), and in the XY plane. And is scanned (scanned) at a scanning speed designated in a predetermined scanning direction (here, the Y-axis direction which is a direction perpendicular to the paper surface in FIG. 1).

  The position along the X-axis and Y-axis directions in the moving plane of reticle stage RST is always detected by reticle interferometer 22 formed of a laser interferometer. Of the reticle interferometers 22, the reticle X interferometer shown in FIG. 1 emits a laser beam directed in the X-axis direction and reflects the laser beam to the X-moving mirror 24 arranged on the reticle stage RST and orthogonal to the X-axis. For example, the position of the reticle stage RST is detected with a resolution of about 0.5 to 1 nm. Although not shown in FIG. 1, a Y moving mirror having a reflecting surface orthogonal to the Y-axis direction is provided on reticle stage RST, and a reticle Y interferometer is provided corresponding to this Y moving mirror. These detect the position of reticle stage RST in the Y-axis direction. Hereinafter, the reticle X interferometer and the reticle Y interferometer are collectively referred to as a reticle interferometer 22.

  Projection optical system PL is arranged below reticle stage RST in FIG. The optical axis AX of the projection optical system PL is oriented in the Z-axis direction orthogonal to the XY plane of the reticle R. The projection optical system PL is a double-sided telecentric reduction optical system, and uses a refractive optical system composed of a plurality of lens elements along the optical axis AX direction. The projection magnification of the projection optical system PL is, for example, 1/5. The slit-shaped illumination area IAR on the reticle R is illuminated by the illumination light IL from the illumination system 12. Illumination light IL that has passed through reticle R on the object plane side is projected as exposure light EL onto wafer W via projection optical system PL, and a reduced image (partially inverted) of reticle R circuit pattern existing in illumination area IAR. Image) is formed in an exposure area IA conjugate to the illumination area IAR on the wafer W on which a photoresist (photosensitive agent) is coated on the surface conjugate with the reticle R on the image plane side.

  Wafer stage WST corresponds to a moving apparatus of the present invention that moves wafer W to a position conjugate with reticle R with respect to projection optical system PL. Wafer stage WST includes an XY stage 14 that is magnetically levitated or gas levitated on the upper surface of stage base 16 and is driven in an XY two-dimensional plane along a stage base 16 by a drive system such as a linear motor or a planar motor. In FIG. 1, the stage that moves in the X direction and the Y direction is integrally shown as an XY stage 14. However, these stages may be two separate layers.

  The wafer table 18 is mounted on the XY stage 14 as a table via a Z / tilt drive mechanism (not shown), and includes a wafer holder 25 as a holding device fixed on the wafer table 18. The wafer W is held by the wafer holder 25 by vacuum suction (or electrostatic suction). The wafer table 18 is finely driven in the three degrees of freedom directions of Z, θx, and θy by a Z / tilt driving mechanism including a voice coil motor and the like.

  As described above, the drive system for the XY stage 14 and the Z / tilt drive mechanism for driving the wafer table 18 are provided separately, but in FIG. 1, these are shown as a wafer drive system 26 for driving them together. ing. In the following description, the XY stage 14 (or wafer stage WST) is driven in the XY two-dimensional plane by the wafer drive system 26, and the wafer table 18 is finely driven in the three-degree-of-freedom directions of Z, θx, and θy. Shall be.

  The wafer table 18 driven in this way is accurately detected by the wafer interferometer 31 on the XY plane, and the signal is sent to the main controller 20. The wafer interferometer 31 includes a wafer X interferometer and a wafer Y interferometer. As shown in FIG. 1, the position along the X-axis direction within the moving surface on the wafer table 18 is always detected by a wafer X interferometer made of a laser interferometer. The wafer X interferometer emits a laser beam directed in the X-axis direction and reflects it to the X-moving mirror 27 arranged on the wafer table 18 and orthogonal to the X-axis, for example, with a resolution of about 0.5 to 1 nm. Thus, the position of the wafer table 18 is detected. Although not shown in FIG. 1, a Y moving mirror having a reflecting surface orthogonal to the Y-axis direction is provided on the wafer table 18, and a wafer Y interferometer is provided corresponding to the Y moving mirror. These detect the position of wafer stage WST in the Y-axis direction.

  The wafer interferometer 31 always detects the position of the wafer table 18 in the X-axis direction and the Y-axis direction, for example, with a resolution of about 0.5 to 1 nm. Specifically, the wafer X interferometer and the wafer Y interferometer are multi-axis interferometers having a plurality of measurement axes. In addition to the X and Y positions of the wafer table 18, rotation (yaw (θz rotation that is rotation around the Z axis) ), Pitching (θx rotation that is rotation around the X axis), and rolling (θy rotation that is rotation around the Y axis)) can also be measured. Therefore, in the following description, it is assumed that the wafer interferometer 31 measures the X, Y, θz, θy, and θx positions of the wafer table 18 in the direction of five degrees of freedom. In addition, the multi-axis interferometer irradiates a laser beam on a reflection surface installed on a gantry (not shown) on which the projection optical system PL is placed via a reflection surface installed on the wafer table 18 with an inclination of 45 °. The relative position information regarding the optical axis AX direction (Z-axis direction) of the projection optical system PL may be detected. For example, the end surface of the wafer table 18 may be mirror-finished to form a reflecting surface (corresponding to the reflecting surface of the X moving mirror 27 and the Y moving mirror).

  The position information (or speed information) of the wafer table 18 measured by the wafer interferometer 31 is sent to the main controller 20, and the main controller 20 passes the wafer drive system 26 based on the position information (or speed information). The position of the XY stage 14 of the wafer stage WST in the XY plane is controlled.

  A reference mark plate FM on which a reference mark for baseline measurement and other reference marks used for baseline measurement of a wafer alignment system (not shown) is formed is fixed on the wafer table 18. The surface of the reference mark plate FM is almost the same height as the wafer W.

  The main controller 20 includes a known computer having a CPU 20a and a storage device 20b, and an interface for inputting / outputting external data to / from these computers. The storage device 20b widely includes, for example, a memory such as a RAM, a hard disk, a storage medium such as a CD-ROM, or a device whose storage is held by communication means. Further, a display device 20c made of an LCD or the like, and an input device 20d made of a keyboard, a mouse or the like are provided. The display device displays an operation interface screen and a plurality of shot areas SHn partitioned on the surface of the wafer W, and displays the shot areas SHn in which an abnormality in flatness is detected to notify the operator. An operating system (OS) for controlling the computer is installed in the hard disk as the storage device 20b, and a process program for controlling exposure and a flatness abnormality detection program for performing the flatness abnormality detection method of the present invention are stored. Yes. In the present embodiment, the flatness abnormality detection program is stored in a state integrated with the process program. The main controller 20 constitutes the controller of the present invention.

  Next, the AF (autofocus) detection system 6 inherently measures the height of the wafer W in the optical axis AX direction (a direction parallel to the Z axis), and based on the measurement result, the main controller 20 performs the measurement. The wafer W of the wafer stage WST is automatically adjusted to a predetermined focal position of the projection optical system PL via the wafer drive system 26. In the present invention, such an AF detection system 6 is used as the detection apparatus of the present invention. The AF detection system 6 of this embodiment is in principle the same as the detection system disclosed in, for example, Japanese Patent Laid-Open No. 56-42205.

  In the irradiation system 60a of the AF detection system 6 of the present embodiment, an illumination light source, an irradiation optical system, a pattern plate on which a slit-shaped opening pattern is formed, and the like constituting the irradiation apparatus are provided. Also, in the light receiving system 60b, a light receiving slit plate in which slits are formed, a focus sensor as a sensor comprising a light receiving element such as a photodiode disposed opposite to each slit of the slit plate, a rotational direction vibration plate , And a light receiving optical system.

  If the accuracy of the wafer W itself is poor, even if the surface of the wafer W is measured, the flatness of the wafer holder 25 cannot be inspected due to variations in the thickness and warpage of the wafer W. Therefore, in the flatness abnormality detection method of the present embodiment, since the abnormality of the flatness of the wafer holder 25 is indirectly inspected by inspecting the surface of the wafer W, normal device manufacture is performed as an object to be detected by the AF detection system 6. A super flat wafer (measuring wafer) is used as a flat substrate set to have a higher flatness than the wafer W used in the above.

  The operation of each part of the AF detection system 6 will be briefly described. When the pattern plate having the slit-shaped opening pattern is illuminated by the illumination light from the illumination light source in the irradiation system 60a, each opening pattern of the pattern plate is changed. The transmitted detection light beam DL is irradiated on the surface of the wafer W through the irradiation optical system. This detection light beam DL is irradiated with a large declination of a predetermined angle, for example, 70 to 80 ° with respect to a direction parallel to the direction of the optical axis AX (that is, the Z axis). At this time, the slit image S projected onto the wafer W is projected so that the longitudinal direction thereof is orthogonal to the detection light beam DL. As a result, an image of a slit-like opening pattern (referred to as “slit image S”) is formed on the surface of the wafer W (see FIG. 2).

  Reflected light from the slit image S formed on the surface of the wafer W is re-imaged on each slit of the light receiving slit plate via the light receiving optical system of the light receiving system 60b, and the light flux of the slit image S is changed. Light is received by the focus sensor.

  In this case, since the light fluxes of the slit images S are vibrated by the rotation direction vibration plate, the position of each re-imaged image on the light receiving slit plate vibrates in a direction perpendicular to the longitudinal direction of each slit. Is done. The sensor detection signal of the focus sensor is synchronously detected by the signal selection processing device 80 with the frequency signal of the rotary diaphragm. At this time, if the slit image S deviates in the height direction from the reference position, the reflected light input to the focus sensor deviates and the amount of light changes. A defocus signal (defocus signal) obtained by synchronously detecting the sensor detection signal by the signal selection processing device 80, for example, an S curve signal, is supplied to the main controller 20 as the height position detection signal Z. ing. Since this position detection signal Z changes in proportion to the difference in the Z coordinate from the reference position, the main control device 20 can calculate the Z coordinate. That is, the slit image S formed on the wafer W becomes a measurement region for measuring the position in the Z direction on the surface of the wafer W.

  The AF detection system 6 uses the best imaging plane (best focus plane) of the projection optical system PL as a reference plane, and irradiates the reference mark plate FM on the wafer table 18 with the detection light beam DL of the irradiation system 60a. Offset adjustment is performed so that the height position detection signal Z becomes 0 when the mark surface of the reference mark plate FM matches the surface. When the slit image S is projected from the irradiation system 60a onto an arbitrary measurement point on the wafer W, the height position detection signal Z is proportional to the difference in height between the reference plane and the measurement point. To change. Further, when the proportional linearity is insufficient, it is preferable to correct so that the difference between the height position detection signal Z and the height has a linear relationship.

  Here, the direction of the slit image S projected on the wafer W will be described. For the purpose of making it less susceptible to the influence of the circuit pattern on the wafer W, the direction along a straight line that intersects the X axis and the Y axis at 45 ° respectively. It is said. Therefore, the detection light beam DL is irradiated from the irradiation system 60a of the present embodiment in a direction along a plane that intersects the XZ plane at 45 ° and also intersects the YZ plane at 45 °. As a result, the slit image S is formed in a direction substantially parallel to the Y = X graph in the coordinate system.

  In the present embodiment, not only the flatness abnormality detection method of the present invention but also high-accuracy leveling is possible. Therefore, the irradiation system 60a for irradiating a large number of imaging light beams and their detected light beams The multi-point focal position detection system includes a light receiving system 60b having a large number of sensors (focus sensors) that individually receive reflected light from the surface of the DL wafer W. Such a multipoint focal position detection system is disclosed in, for example, Japanese Patent Laid-Open No. 6-283403 (corresponding US Pat. No. 5,448,332).

  In the AF detection system 6 of the present invention, in order to detect both inclinations of θx and θy at the same time, the detection is performed from the irradiation system 60a including 49 pattern plates arranged in a matrix of 7 rows × 7 columns. Forty-nine slit images S1 to S49 arranged in a matrix of 7 rows × 7 columns are formed on the wafer W by the light beam DL. These reflected lights are detected by focus sensors arranged in each of the 49 light receiving slit plates of the light receiving system 60b.

  The number of channels that the main control device 20 receives simultaneously is determined according to the number of channels of the signal processing circuit in the signal selection processing device 80. In the leveling of this embodiment, for example, all 49 signals are simultaneously received. Can receive.

  FIG. 2 is a diagram showing a plurality of slit images S1 to S49 irradiated on the wafer W. As shown in FIG. The slit images S1 to S49 are arranged in a matrix of 7 rows × 7 columns, and the measurement center point CP, which is the center point of the center slit image S25, and the center of the exposure area IA (that is, the image plane and light of the projection optical system). (Position where the axis AX intersects). The slit images S1 to S49 are immovable with respect to the exposure area IA. Of these, in the present invention, there are seven slit images S in the middle row among seven rows in the Y-axis direction (first direction), and seven rows aligned in a straight line along the X-axis direction (second direction). Only the slit images S22 to S28 are used. In the present invention, since the Z-direction position of the surface of the wafer W is continuously measured while scanning the wafer W in the Y-axis direction, it is sufficient that at least one slit image S exists at the same X-axis direction position. . On the other hand, all of the plurality of slit images S arranged in the X-axis direction are used to simultaneously inspect a wide area. Details of scanning measurement will be described later.

  In the slit images S22 to S28 formed on the wafer W, the distance between the center points of the slit images S in the X-axis direction is about 3.5 mm. The distance in the X-axis direction between the center of the slit image S22 and the center of the slit image S28 is 21 mm. The width in which one slit image S scans the wafer W, that is, the scanning width D is approximately 1.75 mm, and the gap L between adjacent slit images S is approximately 1.75 mm. That is, in this embodiment, D = L. Thereby, the width of 24.5 mm can be scanned by the seven slit images S22 to S28, and the shot region SHn having a width of 25 mm in the X-axis direction can be substantially covered by one scan.

Next, the flatness abnormality detection method of this embodiment will be described. The flatness abnormality detection method of the present embodiment is roughly divided into the following three procedures. First, the wafer W is scanned, and the AF detection system 6 is used to acquire the position information of the measurement point MP on the wafer W. Next, the surface of the wafer W is divided into a plurality of shot areas SHn according to the shot area at the time of exposure, and an approximate plane acquisition step for obtaining an approximate plane APn for each shot area SHn is performed. Further, in a predetermined case, a corrected approximate plane acquisition step for obtaining a corrected approximate plane APn ′ is performed. Finally, each piece of position information at the measurement point MP included in each shot region SHn is compared with the approximate plane APn or the corrected approximate plane APn ′, and from the approximate plane APn and the corrected approximate plane APn ′. divergence performs steps flatness anomaly detection to extract the shot area SHn as abnormal flatness when greater than a first threshold value K _2.

  In this embodiment, a normal method for detecting a relatively large flatness abnormality at a relatively high speed (hereinafter referred to as “normal mode”) and a high-definition technique for detecting a fine flatness abnormality that takes time. There are two methods (hereinafter referred to as “fine mode”), and these can be arbitrarily used depending on the purpose of inspection. In this embodiment, the wafer W is mounted on the wafer holder 25, but dust may adhere to the surface of the wafer W. Further, before the wafer W is mounted on the wafer holder 25, dust may adhere to the wafer holder 25, and the wafer W may be mounted thereon. In this case, since the wafer W is strongly attracted to the wafer holder 25 by vacuum suction, a part of the wafer W may rise due to the presence of dust. In the latter case, the range showing the flatness abnormality is relatively large and is easily found in the normal mode. On the other hand, the dust on the surface of the wafer W exists from a small thing of about 200 nm, and the fine mode is suitable for finding these.

  Hereinafter, an example of the procedure of the flatness abnormality detection method will be described in detail according to the flowcharts shown in FIGS. The flatness abnormality detection program of this embodiment is integrated with a process program for exposure control of the exposure apparatus 1 and stored in the storage device 20b of the computer of the main control device 20, and the display device 20c of the main control device 20 is stored. Are operated by the input device 20d, the AF detection system 6 and the like are controlled by the CPU 20a, and each process of the flatness abnormality detection method is executed.

  The operator of the exposure apparatus 1 selects a measurement condition file in which various conditions necessary for measurement are recorded prior to the process of detecting an abnormality in flatness, and the main control apparatus 20 reads this (step 1 ( In the flowchart, “step” is abbreviated as “S” and is shown as (S1).)) In the measurement condition file, conditions relating to data sampling, conditions relating to a measurement range, conditions relating to a measurement map, and the like are recorded. In the measurement condition file, the slit image S used for detection, the scanning speed at the time of scanning measurement (relative speed), and the sampling frequency for sampling the detection signal from the focus sensor are recorded as conditions relating to data sampling. In the present embodiment, the slit images S22 to S28 are selected as the slit images S to be detected. The scanning speed is 125 mm / s and the sampling frequency is 125 Hz. Therefore, in terms of distance, sampling is performed every 1 mm along the Y-axis direction.

  In addition, measurement start X coordinate Xs, measurement end X coordinate Xe, scan start Y coordinate Ys, and scan end Y coordinate Ye are recorded as conditions relating to the measurement range. Details of these coordinates will be described later.

  Further, as conditions regarding the measurement map, the wafer size, the size and arrangement of the shot area SHn, and the step pitch are recorded. In the present embodiment, the wafer size is recorded as 8 inches (diameter 200 mm), and the size of each shot area SHn is recorded as a length Lx in the X-axis direction of 25 m and a length Ly in the Y-axis direction of 33 mm, as shown in FIG. As described above, a 5 × 7 array of shot arrays is recorded. The step pitch is determined according to the size of the shot area SHn, and 25 mm is recorded as the X-axis direction step pitch Px and 33 mm is recorded as the Y-axis direction step pitch Py. In the present embodiment, the shot areas SH are all the same and are arranged in a uniform matrix. However, for efficient use of the wafer W, shot areas having different sizes are provided and shots that match the shape are provided. It may be an array of regions.

  The measurement condition file thus selected is read, and the main controller 20 performs control based on each condition described in the measurement condition file. When reading of the measurement condition file is completed, the measurement mode is selected (S2). Here, according to the purpose, either Normal (normal mode) or Fine (fine mode) is input by the operator from the input device 20d such as a keyboard. The CPU 20a of the main control device 20 determines which measurement mode is input, and executes the process of the position information acquisition step in any measurement mode.

  Hereinafter, the step of acquiring the position information when the normal mode is selected (S2; Normal) will be described. First, focusing is performed at the wafer center (S3). The height in the Z direction at the wafer center at this time is set to “0”, which is used as a reference for focusing in other areas thereafter.

  Next, initial setting of the wafer stage position (X, Y) is performed (S4). Here, the measurement start X coordinate Xs and the scan start Y coordinate Ys are set as the initial position of the wafer stage. In the following description, an XY coordinate system set with reference to the stage base 16 is used, and the wafer stage WST is described as moving in this coordinate system. The coordinate system is a mathematical coordinate system with + Y on the paper and + X on the right. Further, the position of the slit image S itself does not change with respect to the XY coordinate system.

  The measurement start X coordinate Xs and the scan start Y coordinate Ys will be described with reference to FIG. FIG. 3 shows a shot map of the wafer W divided into a plurality of shot areas SH. In this embodiment, the wafer W is partitioned into a matrix of 5 rows × 7 columns. Further, the shot center SC that is the center of each shot area SH is represented by a black dot. The measurement start X coordinate Xs is an X coordinate at which the shot center SC at the leftmost end of the wafer (the X coordinate of the wafer is the maximum value) coincides with the measurement center point CP. The scanning start Y coordinate Ys is the Y coordinate of the position where the measurement center point CP is lowered from the lower end Ws of the wafer by a margin Ms of a predetermined distance (here, 10 mm) that is the running distance (the Y coordinate value is large). The reason why the margin Ms is provided is that each shot region SH can be scanned at a constant speed by performing acceleration in the running section. Thus, the wafer stage position is set to (Xs, Ys) as the scanning start position. Then, main controller 20 positions wafer stage WST at (Xs, Ys) via wafer drive system 26 (S5).

  Next, the wafer stage WST set at the initial position (Xs, Ys) is scanned and moved in the −Y direction. Simultaneously, sampling is started at a predetermined sampling frequency (here 125 Hz) (S6). At this time, wafer stage WST accelerates to a predetermined speed (here 125 mm / s) while moving by margin Ms (here 10 mm), and thereafter performs scanning movement at a constant speed of 125 mm / s. In FIG. 3, for convenience of illustration, the slit image S is drawn on the wafer W so as to scan and move in the + Y direction. However, the slit image S does not actually move, and the wafer W becomes the slit image S. In contrast, the scanning movement is performed in the -Y direction.

  By such scanning movement, as shown in FIG. 4, the band-shaped areas SA22 to SA28 of the wafer W are scanned by the slit images S22 to S28. At this time, since main controller 20 samples the detection signal from the focus sensor at 125 Hz, Z-direction position information on the wafer W surface in scanning regions SA22 to SA28 is used as measurement point MP at intervals of 1 mm in the Y-axis direction. It will be measured in.

  The position of wafer stage WST during scanning movement is measured by wafer interferometer 31. Main controller 20 reads the position of wafer stage WST measured by wafer interferometer 31 at the same timing in synchronization with the detection signal from the focus sensor, and stores it in storage device 20b in association with the Z-direction position information. In this way, the Z-direction position information at each measurement point MP is stored together with the XY coordinates of the measurement point MP and accumulated as measurement data. As described above, in the present embodiment, the acquisition of the Z direction position information is controlled by time, so that the position information can be acquired at high speed with a simple configuration.

  Further, the wafer interferometer 31 monitors whether or not the Y coordinate of the wafer stage position (X, Y) has reached the scanning end Y coordinate Ye, and the current Y coordinate and the scanning end Y coordinate Ye. Are compared (S7). If Y = Ye is not satisfied (S7; NO), the scanning movement in the −Y direction is continued (S6). If Y = Ye (S7; YES), the wafer stage WST is driven as the end of scanning. Stop and end scanning (S8). Although not described in the flowchart, if the value of the Y coordinate of the wafer stage position (X, Y) becomes Ye-10 before the end of scanning, the slit image S has passed the wafer upper end We. Therefore, deceleration for stopping wafer stage WST is started.

  Next, the X coordinate of the wafer stage position (X, Y) is compared with the measurement end X coordinate Xe (S9). If X ≦ Xe is not satisfied (S9; NO), the wafer stage WST is moved to the next scanning start position. Re-setting for moving is performed (S10). Here, the X coordinate of the wafer stage position (X, Y) is assumed to be X = X−Px. This means that wafer stage WST is moved in the −X direction by a step pitch Px in the X direction (in FIG. 3, for convenience of illustration, slit images S22 to S28 are moved in the right direction with respect to wafer stage WST ( + X direction)). Further, the Y coordinate of the wafer stage position (X, Y) is set to Y = Ys. This means that wafer stage WST moves in the + Y direction by (Ye−Ys) and returns to the scanning start position. (In FIG. 3, for convenience of illustration, the slit images S22 to S28 are described so as to move downward (−Y direction) with respect to wafer stage WST.) Thereby, the wafer stage position (X, Y) is set to the scanning start position of the second column.

  Based on this setting, the wafer stage position (X, Y) is actually positioned (S5), and thereafter, the wafer stage WST is scanned and moved in the −Y direction as in the first row scanning, and at the same time AF at a predetermined frequency is performed. Sampling (position information detection) is performed (S6). When the Y coordinate becomes the scanning end Y coordinate Ye (S7; YES), the wafer stage is stopped and the scanning is ended (S8), and the X coordinate of the wafer stage position (X, Y) is measured. (S9), if X ≦ Xe is not satisfied (S9; NO), then the third column scan is executed in the same procedure as the second column.

  Thus, scanning is repeated, the X coordinate of the wafer stage position (X, Y) is compared with the measurement end X coordinate Xe (S9), and when X ≦ Xe (S9; YES), the measurement is ended (S9; YES). S11). As described above, by repeating the scanning while changing the position in the X direction, the position information of the wafer W in the Z direction can be measured in a short time without omission. When all measurements are completed, data trimming of measurement data is performed (S12). In this procedure, by referring to the measurement condition file in the sampled data, the exposure effective area of the wafer W excluding the disable range DR which is an unexposed area of about 3 mm provided at the peripheral edge around the wafer W. Data outside the square circumscribing is deleted as unnecessary data. Two opposite sides of this square are parallel to the X axis, and the other two sides are parallel to the Y axis. Then, together with the read measurement condition file (S1) and the identification flag of the selected measurement mode type (normal mode or fine mode), the measured measurement data is associated with one file as a measurement result file in the storage device. Store (S13). This completes the position information acquisition step in the normal mode (END).

  Next, the position information acquisition step when Fine is selected as the measurement mode selection in S2 of FIG. 8 (S2; Fine) will be described with reference to the flowchart shown in FIG. In this procedure, the description overlapping with the procedure shown in FIG. 8 is simplified.

  The position information acquisition step in the fine mode is different from the position information acquisition step in the normal mode in that the scanning in the same shot area SHn is repeated twice. In this case, in the multipoint AF system, if the slit images S formed on the wafer W are in contact with each other or if the adjacent slit images S are too close to each other, they are affected by the reflected light of the adjacent slit images. Therefore, the detection accuracy decreases. For this reason, a predetermined gap is required for the slit image S. In particular, when detecting the tilt of the surface of the wafer W, it is preferable to have a certain gap. In the AF detection system 6 of the present embodiment, the wafer W is leveled by detecting the tilt. On the other hand, it is preferable to inspect the entire surface of the wafer W in order to detect abnormalities in flatness on the wafer W, particularly dust. For this reason, in the present embodiment, in order to achieve both the function of accurately leveling the wafer W and the function of detecting dust in a short time in the same AF system, the same shot area SHn is scanned a plurality of times. By performing (in this embodiment twice), scanning is performed so as to compensate for this gap, so that detection resolution can be increased and detection of dust without leakage can be made.

  As a specific procedure, first, focusing is performed at the wafer center (S23). Next, initial setting of the wafer stage position (X, Y) is performed (S24). The fine mode of the present embodiment is characterized in that the same shot area SHn is scanned twice while shifting the position in order to obtain a resolution twice that of the normal mode. Therefore, in the initial setting of wafer stage position (X, Y), scanning is started from the position where wafer stage WST is moved to the + X side from the normal mode. Here, FIG. 5 is a diagram illustrating the first scanning position in the fine mode. The scan position in the normal mode shown in FIG. 4 is compared with the scan center point CP passing through the shot center SC, and the first scan position in the fine mode shown in FIG. Is different in that scanning is performed so as to pass a position shifted by ½ of the scanning width D to the −X side of the shot center SC. Wafer stage WST is positioned at the wafer stage position at the start of the first scan thus set (S25).

  Next, the measurement center point CP scans (primary scan) the wafer stage WST set at the initial position (Xs + D / 2, Ys) in the −Y direction. At the same time, sampling is performed at a predetermined frequency (125 Hz in this case) (S26). The current Y coordinate and the scanning end Y coordinate are compared (S27), and when Y = Ye (S27; YES), the driving of wafer stage WST is stopped and the primary scanning is ended as the end of scanning (S27). S28).

  Next, the scan start position is reset (S29). Here, it is assumed that X = X−D and Y = Ys. The X coordinate is set to X = X−D, as shown in FIG. 6, in order to prevent the second scanning (secondary scanning) from overlapping the primary scanning area indicated by the two-dot chain line. This is because it is necessary to skip wafer stage WST in the −X direction by the scanning width D from the above area.

Main controller 20 moves and positions wafer stage WST to the set wafer stage position (X, Y) (S30), and performs the second scan movement (secondary) in the Y direction in the same procedure as the primary scan. Scanning) and simultaneously AF sampling is performed (S31).
When Y = Ye (S32; YES), as the end of scanning, the driving of wafer stage WST is stopped and the secondary scanning is ended (S33). As described above, in the fine mode, since one shot area SHn is scanned twice, more detailed measurement data can be obtained as compared with the normal mode. Next, the X coordinate of the wafer stage position (X, Y) is compared with the measurement end X coordinate Xe (S34). If X ≦ Xe−D / 2 is not satisfied (S34; NO), the next scanning start position is reached. Resetting for moving wafer stage WST is performed (S35). Here, the X coordinate of the wafer stage position (X, Y) is set to X = X + D−Px. If the wafer stage WST is in the X-direction step pitch Px normal mode, the step pitch Px may remain as it is. However, in the fine mode, the primary scan is biased to the −X side and the secondary scan is biased to the + X side. This is because the distance + D and the wafer stage WST need to be shifted (returned) to the + X side. The Y coordinate of the wafer stage position (X, Y) is set to Y = Ys. Here, it is set to the scanning start position of the second column.

  Based on this setting, the wafer stage position (X, Y) is actually positioned (S25), and then the second and subsequent scans are performed in the same manner as in the primary and secondary scans of the first row shot area SHn. The procedure of S34 is repeated, and the X coordinate of the wafer stage position (X, Y) is compared with the measurement end X coordinate Xe (S34). When X ≦ Xe−D / 2 is satisfied (S34; YES), the measurement is terminated (S36).

  Then, the process returns to the flowchart of FIG. 8 through the connector 2, and data trimming is performed (S12). The measurement data is measured together with a measurement condition file, an identification flag indicating the type of measurement mode (normal mode or fine mode), and an identification flag for distinguishing primary scanning or secondary scanning in the fine mode. Is stored in the storage device in association with one file as a measurement result file (S13). This completes the position information acquisition step in the fine mode (END).

Next, the procedure of the approximate plane acquisition step and the flatness abnormality detection step in the normal mode will be described with reference to the flowcharts of FIGS.
First, in the position information acquisition step of FIGS. 8 and 9, it is determined which measurement mode “Normal” or “Fine” is selected first (S40). The discrimination is performed by reading the measurement mode identification flag stored in the measurement result file. Here, it is assumed that “Normal” is selected first. When the normal mode is selected (S40; Normal), the measurement data is divided for each shot area (shot area SHn) (S41). Here, n is a shot number of a shot area included in one wafer W, and takes a value of n = 1 to N (N is the total number of shot areas in one wafer).

In the division, the wafer W is divided into a plurality of shot areas SHn based on the shot size setting and shot arrangement setting stored in the measurement condition file, and the measurement data of the measurement point MP recorded in the measurement result file is shot. This is done by grouping for each region SHn. Specifically, the range (X _R ≦ X <X _L , Y _U ≦ Y <Y _D ) of each shot area SHn is compared with the XY coordinates of each measured point MP, and the XY coordinates of the measured point MP are compared. Is within the range of the shot area SHn, it is handled as the measurement point MP of the shot area SHn. Here, _{X L} is the coordinate of the wafer stage WST leftmost shot area SHn coincides with the measurement center point CP, _{X R} are the coordinates of the wafer stage WST right end coincides with the measurement center point CP, _{X L} -X _R = Lx relationship. Further, _{Y D} coordinates of the wafer stage WST the lower end of the shot area SHn coincides with the measurement center point CP, the _{Y U} are the coordinates of the wafer stage WST upper end coincides with the measurement center point CP, _{Y U} -Y _D = It has a Ly relationship. Then, n = 1 is set as an initial value (S42), and the process is started.

  Next, the approximate plane APn corresponding to the shot area SHn is calculated using all of the data Zij that is the measurement data of the measurement point MP included in the shot area SHn (S43). Here, the data Zij represents measurement data (Z direction position information) at the measurement point MP in the i-th row and j-th column among the measurement points MP included in the shot area SHn. At this time, the approximate plane APn is

で、Ｌを最小とする、ａ，ｂ，ｃを求めればよい。Ｌを最小にするための条件は、

Then, a, b, and c that minimize L can be obtained. The conditions for minimizing L are:

であるので、これを計算して書き下し、行列式の形にまとめると

So, if you calculate this and write it down and put it together in the form of a determinant

となる。ここで

It becomes. here

とおけば[数３]の両辺にＡの逆行列をかけて、

If you do, multiply the sides of [Equation 3] by the inverse matrix of A,

と表せるので、これを解くことによりａ，ｂ，ｃを求めることができ、近似平面ＡＰｎが算出される。

By solving this, a, b, and c can be obtained, and the approximate plane APn is calculated.

Here, for all data Zij in the shot area SHn, residual data Z′ij from the approximate plane APn obtained by the above-described procedure is obtained (S44), and the average Z′ave and standard deviation are obtained for the data Z′ij. 3σ _{Z ′} is calculated by the following equation (S45).

続いて、Ｓ４５で求めた標準偏差３σ_Ｚ'と閾値Ｋ_１とを比較する（Ｓ４６）。ここで、標準偏差３σ_Ｚ’が閾値Ｋ_１を超えた場合は（Ｓ４６；ＮＯ）、数値のばらつきが大きすぎて近似平面ＡＰｎによる比較が適当ではないとしてエラーとして判断し、その旨を表示して処理をエラー終了する。一方、標準偏差３σ_Ｚ’が閾値Ｋ_１を超えない場合は（Ｓ４６；ＹＥＳ）、ショット領域ＳＨｎ内の残差のデータＺ’ｉｊとＺ’ａｖｅ±３σ_Ｚ’とを比較する（Ｓ４７）。ここで、Ｚ’ａｖｅ±３σ_Ｚ’を超えるデータＺ’ｉｊがある場合は（Ｓ４８；ＹＥＳ）、当該データに対応する元のデータＺｉｊを除去して、残りのデータＺｉｊを使用して補正後近似平面ＡＰｎ’の算出を行う（Ｓ４９）。

Then, compared with a standard deviation 3 [sigma] _{Z 'obtained} in S45 with the threshold value _{K 1} (S46). Here, when the standard deviation 3σ _{Z ′} exceeds the threshold value K ₁ (S46; NO), it is judged as an error that the comparison of the approximate plane APn is not appropriate because the numerical value variation is too large, and this is displayed. To terminate the process with an error. On the other hand, if the standard deviation 3σ _{Z ′} does not exceed the threshold value K ₁ (S46; YES), the residual data Z′ij in the shot area SHn is compared with Z′ave ± 3σ _{Z ′} (S47). Here, when there is data Z′ij exceeding Z′ave ± 3σ _{Z ′} (S48; YES), the original data Zij corresponding to the data is removed, and after correction using the remaining data Zij The approximate plane APn ′ is calculated (S49).

And it transfers to the procedure shown in the flowchart of FIG. If there is no data Z′ij exceeding Z′ave ± 3σ _{Z ′} (S48; NO), a deviation ΔZij between the approximate plane APn already obtained in S43 and each data Zij is calculated (S50). On the other hand, if there is data Z′ij exceeding Z′ave ± 3σZ _′ (S48; YES), a deviation ΔZij between the corrected approximate plane APn ′ calculated in S49 and each data Zij is calculated (S50). ).

Next, the maximum value Max (ΔZij) and the minimum value Min (ΔZij) are extracted from all the deviations ΔZij included in the shot area SHn. Then, it calculates a difference Max (ΔZij) over Min (ΔZij), comparing determines its value, and a threshold value _{K 2} (S51). This is for determining an abnormality in flatness on the surface of the wafer W. Here, K ₂ is the predetermined threshold value suitable for the determination of the presence of foreign material, and the depth of focus of the projection optical system in which the exposure apparatus is provided is determined based on such a wafer processing process of future exposure. Then, Max If the value of (ΔZij) over Min (ΔZij) exceeds the threshold value _{K 2} (S51; YES), the shot area SHn, determines that there is dust (S51). Further, Max when the value of (ΔZij) over Min (ΔZij) is less than or equal to the value of the threshold value _{K 2} (S51; NO), the shot area SHn is determined that no dust (S53).

  Next, if n = N is not satisfied (S54; NO), the processing of all the shot areas SHn is not completed. Therefore, n = n + 1 is set (S55), and the process returns from the connector 5 to S43 in FIG. Then, the processing of the next shot area SHn is repeated (S43 to S54). Then, when n = N (S54: YES), it is determined that the processing of all the shot areas SHn is completed, and the shot area SHn determined to have dust is displayed on the shot map displayed on the screen of the display device 20c. (S56).

  Here, FIG. 7 is a diagram showing the shot map 21 displayed on the screen of the display device 20c. The shot map 21 displays all the shot areas SHn partitioned on the wafer W. Here, the shot area 21b determined as having dust in S52 is the shot area 21a determined as having no dust in S53. Different correspondences, for example, different colors are displayed. As described above, in the present embodiment, the abnormal data is removed to generate the corrected approximate plane APn ′, and the flatness abnormality is determined based on the corrected approximate plane APn ′. Therefore, this is a reference for determining whether the flatness is abnormal. The virtual plane can be made closer to the real plane, and the flatness abnormality can be determined with higher accuracy. Further, since the approximate plane APn and the corrected approximate plane APn ′ are obtained for each shot region SHn, a local flatness abnormality can also be detected. Further, since the shot area SHn determined to have dust is displayed on the shot map displayed on the screen of the display device 20c, the shot area SHn affected by dust can be recognized at a glance.

Next, the procedure of the approximate plane acquisition step and the flatness abnormality detection step in the fine mode will be described with reference to the flowcharts of FIGS.
First, in the position information acquisition step of FIGS. 8 and 9, it is determined by the identification flag stored in the measurement result file whether the measurement mode “Normal” or “Fine” is selected first (S40). ). Here, it is assumed that “Fine” is selected. A description of the same procedure as that in the normal mode in FIGS. 10 and 11 is omitted. In the flowchart shown in FIG. 10, when Fine is selected as the measurement mode (S40; Fine), the process proceeds to the procedure of S60 in FIG. Here, the measurement data is divided for each shot area, and is divided into N shot areas SHn, where n = 1 to N. Then, n in the shot area SHn is initialized as n = 1 (S61).

Next, in the fine mode, the primary scan and the secondary scan are performed twice. Since it is necessary to distinguish between these, the primary scan or the secondary scan stored in the measurement result file is used. The identification flag is read and determined (S62). Here, a case where data processing is performed on a primary scan file (S62; YES) will be described first. First, the approximate plane AP ₁ n is calculated using all the primary scanning measurement data Z ₁ ij included in the shot area SHn (S63). Next, the data _{Z 1} of all primary scanning measurement, 'seek ij (S64), the data _{Z 1'} data _{Z 1} of the residuals from the approximate plane AP ₁ n obtained average _{Z 1} for ij 'ave, standard Deviation 3σ _{Z1 ′} is calculated (S65). Then, it is determined whether or not 3σ _{Z1 ′} <K ₁ (S66). If 3σ _{Z1 ′} <K _{1 is} not satisfied (S66; NO), the process ends as an error. Also, _'if> a _{K 1} (S66; YES), the data _{Z 1'} 3 [sigma] _Z1 Compare ij and _{Z 1 'ave ± 3σ Z1'} (S67). Then, it is determined whether or not there is data exceeding Z ₁ 'ave ± 3σ _Z1' (S68). If there is data exceeding Z ₁ 'ave ± 3σ _Z1' (S68; YES), it corresponds to the data. The original data Z ₁ ij to be removed is removed, and the corrected approximate plane AP ₁ n ′ is calculated using the remaining data Z ₁ ij (S69).

Next, when there is no data Z ₁ 'ij exceeding Z ₁ ' ave ± 3σ _{Z1 '} (S68; NO), a deviation ΔZ ₁ ij between the approximate plane AP ₁ n calculated in S63 and each Z ₁ ij is calculated. (S70). On the other hand, when there is data Z ₁ 'ij exceeding Z ₁ ' ave ± 3σ _{Z1 '} (S68; YES), the deviation ΔZ between the corrected approximate plane AP ₁ n' calculated in S69 and each data Z ₁ ij ₁ ij is calculated (S70). Subsequently, the maximum value Max (ΔZ ₁ ij) and the minimum value Min (ΔZ ₁ ij) are extracted from all the deviations ΔZ ₁ ij included in the shot region SHn. Then, the difference Δ1 = Max (ΔZ ₁ ij) −Min (ΔZ ₁ ij) is calculated (S71). Up to this point, the primary scan data processing is completed, and the process proceeds to the secondary scan data processing from S72 of the flowchart shown in FIG.

Here, similarly to the primary scanning, first, the approximate plane AP ₂ n is calculated using all the secondary scanning measurement data Z ₂ ij included in the shot region SHn (S72). Next, with the data _Z 2 ij of all secondary scanning measurement, 'seek ij (S73), the data _{Z 2'} data _{Z 2} of residuals from the approximate plane AP ₂ n calculated for ij average _{Z 2} 'ave Then, the standard deviation 3σ _{Z2 ′} is calculated (S74). Then, it is determined whether or not 3σ _{Z2 ′} <K ₁ (S75), and if 3σ _{Z2 ′} <K ₁ , the process ends as an error. Furthermore, 3 [sigma] _{Z2 '<if K 1,} data _{Z 2'} compares the ij and _{Z 2 'ave ± 3σ Z2'} (S76). Then, it is determined whether or not there is data exceeding Z ₂ 'ave ± 3σ _Z2' (S77). If there is data exceeding Z ₂ 'ave ± 3σ _Z2' (S77; YES), it corresponds to the data. The original data Z ₂ ij to be corrected is removed, and the corrected approximate plane AP ₂ n ′ is calculated using the remaining data Z ₂ ij (S78).

Next, when there is no data Z ₂ 'ij exceeding Z ₂ ' ave ± 3σ _{Z2 '} (S77; NO), the deviation ΔZ ₂ ij between the approximate plane AP ₂ n calculated in S72 and each data Z ₂ ij is calculated. Calculate (S79). On the other hand, when there is data Z ₂ 'ij exceeding Z ₂ ' ave ± 3σ _{Z2 '} (S77; YES), the deviation ΔZ ₂ between the corrected approximate plane AP ₂ n' calculated in S78 and each Z ₂ ij ij is calculated (S79). Subsequently, the maximum value Max (ΔZ ₂ ij) and the minimum value Min (ΔZ ₂ ij) are extracted from all ΔZ ₂ ij included in the shot area SHn. Then, the difference Δ2 = Max (ΔZ ₂ ij) −Min (ΔZ ₂ ij) is calculated (S80). As described above, the primary scanning and the secondary scanning are performed separately and independently.

Here, since the calculation of Δ1 and Δ2 is completed, the larger one of Δ1 and Δ2 (Max (Δ1, Δ2)) and K ₂ are compared, and Max (Δ1, Δ2)> K ₂ or not. Is determined (S81). Then, Max (Δ1, Δ2)> when a _{K 2} (S81; YES), the the shot area SHn, it is determined that there is dust (S83). Here, it is not distinguished whether dust is found in the primary scan or dust is found in the secondary scan. On the other hand, Max (Δ1, Δ2) when a _{≦ K 2 (S81; NO)} , the the shot area SHn is determined that there is no dust. In this case, no dust was found in either the primary scan or the secondary scan. Since the determination of this shot area SHn is completed, it is determined whether n = N (S84). If n ≠ N (S84; NO), there is a next shot area SHn to be determined. Therefore, n = n + 1 is set, and the process returns to S62 in FIG. Repeat the procedure. If n = N is determined in S84 (S84; YES), the processing of all the shot areas SHn is completed, so that the shot area SHn determined to have dust is displayed on the shot map (S86). The process is terminated. As described above, in the fine mode, the determination of the abnormality in flatness is performed independently based on the measurement data measured by a plurality of independent (here, twice) scans. An abnormality in flatness can be accurately detected.

  Next, an exposure operation when the pattern of the reticle R is sequentially transferred to each shot area SHn on the wafer W by the exposure apparatus 1 of the present embodiment will be briefly described. In this case, as a premise, dust on the wafer is inspected by the flatness abnormality detection method as described above, and if a flatness abnormality is found, the cause of the flatness abnormality such as dust is investigated, and It is assumed that the cause of the abnormality in flatness has been removed. As a premise, a reticle alignment system (not shown), a reticle alignment using a reference mark plate FM on the wafer table 18 and a wafer alignment system (not shown), baseline measurement of the wafer alignment system, and wafer alignment (EGA, etc.) The preparatory work such as has been completed. Note that the above-described preparation operations such as reticle alignment and baseline measurement are disclosed in detail in, for example, Japanese Patent Laid-Open No. 4-324923 (corresponding US Pat. No. 5,243,195), and the following EGA is disclosed in Japanese Patent Laid-Open No. This is disclosed in detail in Japanese Laid-Open Patent Publication No. 61-44429 (corresponding US Pat. No. 4,780,617) and the like, and since it is publicly known, detailed description thereof is omitted.

  First, main controller 20 drives XY stage 14 via wafer drive system 26 based on the result of wafer alignment, and starts an acceleration start position (scan start) for exposure of first shot region SH1 on wafer W. The wafer table 18 is moved to (position).

  Next, main controller 20 starts relative scanning in the Y-axis direction between reticle R and wafer W, that is, reticle stage RST and wafer stage WST. When reticle stage RST and wafer stage WST reach their respective target scanning speeds and reach a constant speed synchronization state, the pattern area of reticle R starts to be illuminated by illumination light IL from illumination system 12, and scanning exposure is started. The relative scanning is performed by the main controller 20 controlling the reticle drive system and the wafer drive system 26 (not shown) while monitoring the measurement values of the wafer interferometer 31 and the reticle interferometer 22 described above.

  Main controller 20 has a projection magnification (1/4) of projection optical system PL, particularly during the above-described scanning exposure, in which the movement speed Vr of reticle stage RST in the Y direction and the movement speed Vw of wafer stage WST in the Y-axis direction are Alternatively, the synchronization control is performed so that the speed ratio is maintained according to 1/5 times.

  Then, different areas of the pattern area of the reticle R are sequentially illuminated with the illumination light IL, and the illumination of the entire pattern area is completed, thereby completing the scanning exposure of the first shot area SH1 on the wafer W. As a result, the pattern of the reticle R is reduced and transferred to the first shot region SH1 via the projection optical system PL.

  During the scanning exposure described above, the main controller 20 performs so-called pre-reading control based on a defocus signal from a focus sensor located in front of the exposure area IA of the multipoint AF system (60a, 60b) in the scanning direction. The wafer table 18 is positioned near the image plane via the wafer drive system 26. At the same time, based on the defocus signal from the focus sensor in the exposure area IA, the Z position of the wafer table 18 is set via the wafer drive system 26 so that the surface of the wafer W matches the depth of focus of the image plane. And the tilt (pitching amount (θx rotation amount) and rolling amount (θy rotation amount)) with respect to the XY plane are controlled. That is, the auto focus / auto leveling control of the wafer W is performed in this way.

  As described above, when the scanning exposure of the first shot area SH1 is completed, the main controller 20 moves the wafer stage WST stepwise in the X and Y axis directions via the wafer drive system 26, and the second shot area SH2. Is moved to the scanning start position (acceleration start position) for the exposure of the first.

Then, the operation of each part is controlled by the main controller 20 as described above, and the same scanning exposure as described above is performed on the second shot region SH2 on the wafer W.
In this way, the scanning exposure of the shot area SHn on the wafer W and the stepping operation between the shot areas SHn are repeatedly performed, and the pattern of the reticle R is sequentially transferred to all of the shot areas SHn to be exposed on the wafer W. The

When the pattern transfer to all exposure target shots on the wafer W is completed, the next wafer W is replaced, and the alignment and exposure operations are repeated as described above.
The flatness abnormality detection method of the present invention is performed at the time of exposure of the first wafer W of the product lot or at any other time. Therefore, various troubles caused by an abnormality in flatness such as dust are prevented in advance, and the throughput of the product is not lowered.

  As is clear from the above description, in the present embodiment, the main control device 20 constitutes a control device and a calculation device. However, these devices may be configured as separate devices. Further, the flatness measuring apparatus is configured including the main controller 20, the wafer table 18, the wafer drive system 26, the multipoint AF system (60a, 60b), and the like.

The embodiment configured as described above has the following effects.
(1) In this embodiment, since the height information is detected while the wafer W and the slit image S are relatively moved by the wafer stage WST, the position information can be acquired continuously at a relatively high speed. In addition, since the shot area SHn having an abnormality in flatness is automatically extracted and displayed on the display device 20c, there is an effect that troubles caused by dust and the like can be reduced extremely easily and quickly.

  (2) Further, in the present embodiment, after generating the approximate plane APn, a value that deviates by a predetermined value or more is removed, and the corrected approximate plane APn ′ is regenerated to determine the flatness abnormality. There is an effect that a high determination can be made. In addition, the wafer W can be divided with the same shot size and shot arrangement as the exposure process is performed, and the flatness abnormality can be determined for each shot area, so that it conforms to the actual exposure process. There is an effect that the flatness inspection of the wafer W can be performed.

(3) Since the acquisition of the position information is controlled by time, there is an effect that the position information can be acquired with a simple configuration as compared with the case where the control is performed at the moving position.
(4) By making use of the function of the wafer stage WST of the exposure apparatus, the position information of the wafer W can be acquired in a short time by the operation by scanning and step, and the inspection time can be shortened. Yes.

(5) In particular, by projecting a plurality of slit images S and using a plurality of sensors simultaneously, there is an effect that position information of a large number of points can be acquired in a short time.
(6) In the fine mode, the configuration of the AF detection system 6 is used to improve the resolution by control without a new configuration.

  (7) Further, in the fine mode, since the inspection result in the shot area by the primary scanning and the inspection result in the shot area by the secondary scanning are processed independently, there is no need to adjust both data and simple control is performed. In addition, the detection of dust can be ensured by scanning with less leakage.

  (8) In this embodiment, when it is determined that there is an abnormality in flatness, the shot area is shown on the display device, so that even if there is no detailed analysis or no prior knowledge, the presence of dust immediately Can be notified.

In addition, this embodiment can be implemented as follows.
O One approximate plane APn of a plurality of shot areas SHn is compared with an approximate plane APn of another shot area SHn different from this. When the difference between the approximate plane APn of the one shot area and the approximate plane APn of the other shot area is larger than a predetermined threshold, the inspection area is extracted as an abnormality in flatness in the one inspection area. An approximate plane abnormality detection step may be further included.

  With this configuration, even when it is determined that there is no abnormality in the flatness in the approximate plane in one shot area SHn, the height is greatly different from the approximate plane in other shot areas. This inspection area can be extracted with this as an abnormality in flatness. For example, there is an effect that even when the entire inspection area is lifted up like dust between the wafer and the wafer holder, this can be detected.

○ first threshold K ₂ is compared with a standard deviation 3 [sigma], for example may be compared with the standard deviation 6 [sigma. Further, other threshold values can be arbitrarily set.
The determination of the deviation is based on the distance in the optical axis direction (Z-axis direction), but may be a distance from the approximate plane APn and the corrected approximate plane APn ′, for example.

The method of obtaining the approximate plane APn and the corrected approximate plane APn ′ is based on the least square method, but various calculation methods can be employed in addition to the arithmetic mean.
As an example of the display device 20c, an LCD is used. However, an inspection region having an abnormality in flatness may be displayed by a printing device, and may be displayed by characters regardless of an image as long as it is a figure. it can.

The display mode is not limited to the display of the shot area SHn where there is an abnormality, but also the location where the flatness abnormality is present in the shot area SHn, for example, the maximum value (Max (ΔZij), Max (ΔZ ₁ ij), Max The position corresponding to (ΔZ ₂ ij)) and the minimum value (Min (ΔZij), Min (ΔZ ₁ ij), Min (ΔZ ₂ ij)) may be displayed. According to this, the state of flatness can be grasped more accurately.

  In the fine mode of the present embodiment, the scanning width D and the gap L are the same, and the same shot area SHn is scanned over the entire area by two scans of the primary scan and the secondary scan. Depending on the configuration of the slit image formed on W, the same shot area SHn may be scanned three or more times.

  In addition, a plurality of scans in the fine mode may be performed so as to overlap each other. This is because even if they overlap, since the abnormality in flatness is determined independently of each other, no problem arises. Rather, if one of them cannot detect dust under some condition, it can be detected within the overlapping range.

In the fine mode, a plurality of scans are performed independently of each other, but may be performed in association with each other.
In this case, in this embodiment, AF sampling is performed at a constant interval by scanning at a constant speed. However, the measurement point MP is determined by feeding back the position on the XY plane by the wafer interferometer 31. Also good.

In addition, although the approximate plane APn of the embodiment generates the corrected approximate plane APn ′, this procedure may be omitted.
The flowchart shown in the embodiment is an example, and the procedure can be added, deleted, or changed.

In the present embodiment, the exposure apparatus 1 has been described as an example, but the present invention can also be implemented in, for example, a wafer W inspection machine.
Note that the present invention is not limited to the embodiments, and it goes without saying that those skilled in the art can make improvements and modifications without departing from the scope of the claims.

1 is a diagram showing an outline of an exposure apparatus according to an embodiment. The figure which shows the some slit image irradiated on the wafer. It is a figure which shows the relative movement of the slit image with respect to a wafer. The figure explaining the scan in a normal mode. The figure which shows the primary scanning in a fine mode. The figure which shows the secondary scanning in a fine mode. The figure which shows the shot map displayed on the screen of a display apparatus. The flowchart which shows the procedure of the flatness abnormality detection method. The flowchart which shows the procedure of the flatness abnormality detection method. The flowchart which shows the procedure of the flatness abnormality detection method. The flowchart which shows the procedure of the flatness abnormality detection method. The flowchart which shows the procedure of the flatness abnormality detection method. The flowchart which shows the procedure of the flatness abnormality detection method.

Explanation of symbols

APn (AP ₁ n, AP ₂ n) ... approximate plane, APn '(AP ₁ n', AP ₂ n ') ... corrected post-correction plane, AX ... optical axis, D ... scanning width, DL ... detected light flux, K ₁ ... second threshold, K ₂ ... first threshold, L ... gap, MP ... measured point, PL ... projection optical system, S (S22 to S28) ... slit image, SHn ... shot region, W ... wafer, WST ... wafer stage, Xs ... measurement start X coordinate, Xe ... measurement end X coordinate, Ys ... scan start Y coordinate, Ye ... scan end Y coordinate, Zij, Z ₁ ij, Z ₂ ij ... (by scanning measurement) data, i ... column number (slit number, sensor number), j ... row number, 1 ... exposure apparatus, 6 ... AF detection system, 12 ... illumination system, 14 ... XY stage, 18 ... wafer table, 60a ... irradiation system, 60b ... light reception System, 20 ... main control device, 20c ... display device.

Claims (10)

A flatness abnormality detection method for detecting position information in the height direction of the surface of the inspection object in a plurality of measurement areas set on the surface of the inspection object, and detecting an abnormality in flatness of the surface of the inspection object. ,
A step of acquiring position information by detecting the position information while relatively moving the object to be inspected and the measurement area, and obtaining the position information at a plurality of points to be measured on the surface of the object to be inspected;
Approximating the surface of the object to be inspected into a plurality of inspection areas and obtaining an approximate plane of the inspection area from the position information of the plurality of measurement points included in one inspection area of the plurality of inspection areas A plane acquisition step;
Each of the position information at the plurality of measurement points included in the one inspection region is compared with the approximate plane, and the deviation from the approximate plane is compared with a first threshold value, and the deviation is the first threshold. A flatness abnormality detection method comprising: a flatness abnormality detection step of extracting the inspection region as an abnormality of flatness when larger than a threshold value. In the approximate plane obtained in the approximate plane acquisition step, the deviation of the position information at each measured point from the approximate plane is compared with a second threshold value, and the position information data larger than the second threshold value is obtained. And further comprising a step of obtaining a corrected approximate plane that obtains a corrected approximate plane corrected from the position information of the remaining measured point,
The flatness abnormality detection method according to claim 1, wherein the flatness abnormality detection step performs the flatness abnormality detection step using the corrected approximate plane as an approximate plane. In the position information acquisition step, the position information at the measurement point is detected by detecting the position information at regular time intervals while relatively moving the object to be inspected and the measurement region along the first direction at a constant relative speed. 3. The flatness abnormality detection method according to claim 1, wherein information is obtained. In the position information acquisition step, the relative movement is performed in the second direction perpendicular to the first direction with respect to the relative position between the object to be inspected and the measurement region. 4. The flatness abnormality detection method according to claim 3, wherein a plurality of repetitions are performed at different positions. The flatness abnormality detection method according to claim 4, wherein the plurality of measurement regions are arranged at predetermined intervals along the second direction on the surface of the object to be inspected. The step of acquiring the position information performs a plurality of the relative movements with respect to the same inspection region, and the step of acquiring the approximate plane is independent for each of the position information detected by each of the relative movements. Do,
6. The flatness abnormality step includes extracting the inspection area as having an abnormal flatness when the flatness is determined to be abnormal in any one of the relative movements. The flatness abnormality detection method according to any one of the above. Compared with one approximate plane of the plurality of inspection areas and an approximate plane of another inspection area different from the one approximate plane, the difference between the one approximate plane and the other approximate plane is third. 7. The method according to claim 1, further comprising an approximate plane abnormality detection step of extracting the inspection area as having an abnormality in flatness when the inspection area is larger than the threshold value. The flatness abnormality detection method according to claim 1. In addition to displaying a diagram of a plurality of inspection areas partitioned on the surface of the object to be inspected on the display device, the inspection area extracted as having an abnormal flatness in the flatness abnormality detecting step, and an abnormality in the flatness The flatness abnormality detection method according to any one of claims 1 to 7, wherein the inspection area that has not been detected is displayed in a different manner. A detection device configured to form a plurality of measurement regions by irradiating a detection light beam on the surface of the object to be inspected, and to detect position information in a height direction of the surface of the object to be inspected in the measurement region;
A moving device for relatively moving the object to be inspected and the detection device;
An exposure apparatus comprising: a control device that controls the moving device and the detection device to detect an abnormality in flatness of the surface of the object to be inspected;
The control device includes:
Position information acquisition means for detecting the position information during relative movement between the object to be inspected and the measurement area and obtaining the position information at a plurality of points to be measured on the surface of the object to be inspected;
Approximating the surface of the object to be inspected into a plurality of inspection areas and obtaining an approximate plane of the inspection area from the position information of the plurality of measurement points included in one inspection area of the plurality of inspection areas Plane acquisition means;
Each of the position information at the plurality of measurement points included in the one inspection region is compared with the approximate plane, and the deviation from the approximate plane is compared with a first threshold value, and the deviation is the first threshold. An exposure apparatus comprising: flatness abnormality detection means for extracting the inspection region as an abnormality in flatness when larger than a threshold value. A diagram of a plurality of inspection areas partitioned on the surface of the object to be inspected is displayed, and the flatness abnormality detecting means has detected no abnormality in flatness in the inspection area extracted as having an abnormal flatness. The exposure apparatus according to claim 9, further comprising a display device that displays in a manner different from the inspection area.

Images