JP4251296B2 - Measurement method, adjustment method, marked object, and detection apparatus - Google Patents

Description

  The present invention relates to a measurement technique and an adjustment technique of a detection device for detecting the position of a test mark by receiving a light beam from the test mark, for example, for example, a semiconductor integrated circuit, an image sensor (CCD, etc.), An alignment sensor provided in an exposure apparatus used in a lithography process for forming a fine pattern such as a liquid crystal display or a thin film magnetic head, or an overlay error measurement for measuring an overlay error between a plurality of layers on a substrate. It is suitable for use when adjusting a device or the like.

  Projection exposure apparatus for transferring an image of a reticle pattern as a mask to each shot area on a wafer (or glass plate or the like) coated with a photoresist via a projection optical system when manufacturing a semiconductor integrated circuit or the like (Steppers etc.) are used. For example, a semiconductor integrated circuit is formed by stacking circuit patterns of several tens of layers on a wafer in a predetermined positional relationship. For example, when projecting and exposing circuit patterns of the second and subsequent layers onto a wafer, It is necessary to perform alignment (alignment) between the circuit pattern (existing pattern) formed by the process so far in each shot area and the image of the reticle pattern to be exposed from now on with high accuracy. For this reason, the projection exposure apparatus is provided with an alignment sensor for detecting the position of an alignment mark (wafer mark) attached to each shot area on the wafer together with a circuit pattern.

There are various types of alignment sensors. Recently, an imaging method (image processing method), which is hardly affected by the asymmetry of the wafer mark, is becoming mainstream. This is a system that includes an optical system having the same configuration as a microscope, picks up an image of a wafer mark enlarged by an objective lens, and detects the position of the wafer mark from the image signal.
In addition, an overlay error measurement device (registration measurement device) is used to confirm the overlay accuracy of a pattern exposed by projection exposure with an existing pattern. The position detection device provided in the overlay error measurement device is also an optical system similar to the imaging type alignment sensor provided in the exposure device, but the measurement target is at the position (absolute position) of a single wafer mark. Rather, the relative positional deviation amount between the lower layer mark (existing mark) and the upper layer mark (new mark) is obtained.

  An optical characteristic error in the alignment sensor or the optical system of the position detection device in the overlay error measuring device, that is, an aberration (coma aberration, etc.) of a detection optical system such as an imaging system, or an adjustment error of an illumination system If (such as a displacement of the illumination system aperture stop) remains, an error occurs in the position detection value due to an error in its optical characteristics. This error is generally referred to as TIS (Tool Induced Shift) in the sense of an error caused by the apparatus.

  Recently, as an adjustment method of the optical system of the position detection device for reducing the TIS, two kinds of uneven marks (step marks) having different step amounts are provided close to each other on the evaluation substrate. In addition, a method of adjusting the optical system based on the measured value of the interval between these marks (hereinafter referred to as “different step mark method”) has been proposed. This different step mark method is disclosed in, for example, T. Kanda, K. Mishima, E. Murakami and H. Ina: Proc. SPIE, Vol. 3051, pp.846-855 (1997). In this method, after measuring the distance D1 between the concave and convex marks having two different steps on the wafer, the wafer is rotated 180 ° and the distance D2 between the two concave and convex marks is measured again. In this case, ½ of the difference between the measurement value at the rotation angle of 0 ° and the measurement value at the rotation angle of 180 °, that is, (D1−D2) / 2 is TIS, and this TIS falls within the allowable range. In addition, the optical system is adjusted.

  In many cases, the overlay error measuring apparatus measures a box-in-box mark composed of an outer frame mark and an inner frame mark. Therefore, the two-dimensional positional deviation amount of the center of the inner frame mark measured with respect to the center of the outer frame mark on the evaluation substrate is set to (ΔX1, ΔY1), and the wafer is rotated 180 °. Assuming that the two-dimensional positional deviation amount at the center of both marks obtained by measurement again is (ΔX2, ΔY2), the TIS (Ta, Tb) of the overlay error measuring apparatus is ((ΔX1 + ΔX2) / 2, (ΔY1 + ΔY2) / 2). Also in this case, the optical system is adjusted so that (Ta, Tb) as the TIS falls within the allowable range.

As described above, conventionally, a different step mark method has been proposed in order to correct TIS which is an error caused by the position detecting device. However, this different step mark method has a disadvantage in that it is difficult to accurately form two kinds of concave and convex marks close to each other in a state where the step amounts are different from each other and each set to a predetermined step. .
Even if the uneven marks with different steps can be accurately formed, the different step mark method is effective for adjusting the positional deviation of the illumination system aperture stop. Adjustment may not always be performed with high accuracy.

  Further, conventionally, in order to correct the TIS, after measuring the distance between the pair of marks for evaluation on a predetermined substrate or the amount of relative displacement, the substrate is rotated by 180 ° and the pair of marks again. Since the TIS is obtained by measuring the interval or the amount of relative displacement, there is a disadvantage that it takes time for the measurement. Further, usually, after obtaining the TIS and adjusting the predetermined optical member, the measurement is performed by rotating the substrate by 180 ° until the TIS actually falls within the allowable range, Since it is necessary to repeat the adjustment operation of the optical member, the time required for measurement and adjustment becomes very long, and the measurement error remains when the rotation angle of the substrate cannot be accurately set to 180 °. was there.

  Further, providing a rotating stage capable of rotating the substrate by 180 ° on the stage on which the substrate is placed is not practical because the structure of the stage is complicated and the size is increased. Therefore, after measuring the distance between a pair of marks on the substrate on the stage, once removing the substrate from the stage, rotating the substrate by 180 °, and placing it on the stage again. Then, there is a possibility that foreign substances may adhere to the substrate, and the attaching / detaching operation of the substrate is complicated.

  Furthermore, conventionally, after detecting the interval between two marks having different steps, the marks are rotated by 180 ° and the interval is detected again, and the difference between the two intervals thus detected is reduced to 1/2. It was TIS. This means that the average value of the detection results of the two intervals is regarded as the reference value (true value) of the interval between the two marks at the different steps. However, when two marks having different steps are rotated by 180 °, the overall shape of the mark changes, and errors such as distortion other than TIS may be mixed in the interval detection result.

In view of the above, it is a first object of the present invention to provide a measurement technique and an adjustment technique for a detection apparatus that can easily and accurately form marks for characteristic measurement.
The present invention further provides an adjustment technique for a detection apparatus that can accurately form a mark for characteristic measurement and can accurately correct a predetermined aberration of a detection optical system or an adjustment residual of an illumination system. This is the second purpose.

A third object of the present invention is to provide an optical system adjustment technique capable of measuring an error (TIS) caused by the apparatus in a short time and with high accuracy.
Furthermore, a fourth object of the present invention is to provide a marked object that can be used when the above adjustment technique is performed.
In addition, a fifth object of the present invention is to provide a detection device adjustment technique capable of adjusting an error (TIS) caused by the device with high accuracy.

  Furthermore, a sixth object of the present invention is to provide a detection device or a pattern detection device that can use such an adjustment technique.

A measurement method according to the present invention is a measurement method for measuring an aberration of a detection optical system of a detection device that detects a mark via a detection optical system, and includes a first pattern unit including a pattern having a first width; The illumination light is irradiated to the test mark in which the second pattern portion including the pattern having the second width narrower than the first width is arranged adjacent to the measurement direction, and the illumination light is illuminated with the illumination light. The light flux from the detected mark is condensed on a predetermined surface via the detection optical system, and an image of the detected mark is formed on the predetermined surface, and the image is formed on the predetermined surface. The distance information between the first pattern portion and the second pattern portion of the test mark is measured based on the image of the test mark, and the detection optical system is measured based on the measurement result of the test mark. The predetermined optical characteristic is measured.

Since the test mark for measuring characteristics according to the present invention can be easily and accurately formed, the optical characteristics can be measured accurately.
Moreover, the adjustment method according to the present invention is based on the coma aberration or astigmatism measured using the measurement method of the present invention when the predetermined optical characteristic is coma aberration or astigmatism, for example. The detection optical system is adjusted. According to the present invention, for example, an error caused by a device such as a detection optical system can be adjusted with high accuracy in a short time.

The mark object according to the present invention is a mark object provided with a test mark for measuring characteristics of a detection device that detects the mark based on a light beam generated from an illuminated mark, and the test mark is A first pattern portion including a pattern having a first width and a second pattern portion including a pattern having a second width narrower than the first width, and the first and second pattern portions are In order to measure the interval information, they are arranged adjacent to each other in the measurement direction. The present invention can be used in the measurement method or adjustment method of the present invention.

  The detection apparatus according to the present invention is a detection apparatus that detects a mark via a detection optical system, and includes a first pattern portion including a pattern having a first width, and a second pattern narrower than the first width. A test mark in which a second pattern portion including a pattern having a width of 2 is adjacently arranged in the measurement direction is imaged through the detection optical system, and predetermined optical characteristics of the detection optical system are based on the imaging result Has been adjusted. In the present invention, the adjustment method of the present invention is used.

  The following inventions are also described in the embodiments of the present invention. In other words, the adjustment method of the first position detection device described in the embodiment of the invention of the present application is a detection optical system (10, 9,...) That collects light beams from one or a plurality of test marks. 12, 15, 16, 21), and a position for detecting the position of one test mark or the relative positions of the plurality of test marks based on the light beam collected by the detection optical system. A method for adjusting a detecting device, wherein concave portions (31a, 32a) and convex portions (31b, 32b) are alternately and periodically arranged in a predetermined measurement direction on a predetermined substrate (11), and A plurality of grid marks (DM1, DM2) having different ratios between the widths of the recesses and the projections are formed in close proximity, and the plurality of grid marks are formed via the detection optical system. The distance between the marks (DM1, DM2) in the measurement direction ( d) measuring a and adjusts the predetermined optical characteristics of the optical system for the detection on the basis of the measured value.

  According to the present invention, as the characteristic measurement mark, for example, the first lattice mark (DM1) in which the concave portion (31a) having a width a and the convex portion (31b) having a width b are periodically arranged, A second lattice mark (DM2) in which concave portions (32a) having a width c and convex portions (32b) having a width d are periodically arranged is used. At this time, the ratio (a: b or a / b) between the width of the concave portion and the width of the convex portion of the first lattice mark (DM1) is the width of the concave portion of the second lattice mark (DM2). And the ratio of the width of the convex portion (c: d or c / d) is different. The duty ratio for one pitch of the recess (31a) is 100 × a / (a + b) (%), and the duty ratio for one pitch of the recess (32a) is 100 × c / (c + d) (%). The duty ratio is also different. According to the present invention, the steps of the plurality of grid-like marks may be substantially the same, and only the ratio between the width of the concave portion and the width of the convex portion needs to be different. By using a mask on which a plurality of original patterns are formed, these lattice marks can be formed easily and accurately in a normal lithography process.

  As described above, the present invention uses a plurality of lattice marks having different duty ratios expressed as a percentage of the ratio between the width of the concave portion and the width of the convex portion, and in turn the ratio of the width of the concave portion (or the convex portion) to one pitch. Therefore, the adjustment method can be called “different ratio mark method”. In this case, if asymmetric aberration such as coma remains in the detection optical system, the position of each mark image is shifted according to the duty ratio. Accordingly, when the interval between the mark images is measured, the measured value of the interval between the mark images is deviated from the reference value (design value or the like) in the state where the asymmetric aberration remains, and is the same as the reference value in the state where there is no asymmetric aberration. By utilizing this and adjusting the optical characteristics so that the measured value becomes the reference value while measuring the interval between the mark images, it is possible to easily drive the asymmetric aberration to the allowable range.

Further, it is desirable that the ratio of the width of the concave portion and the width of the convex portion of one of the plurality of lattice marks (DM1, DM2) is 1: 1. The image of the mark having a ratio of the width of the concave portion to the width of the convex portion of 1: 1 can be used as a reference mark when comparing the intervals because the lateral shift hardly occurs due to asymmetric aberration.
The detection optical system is, for example, an imaging optical system that projects the images of the plurality of lattice marks onto a predetermined observation surface, and the optical characteristics of the adjustment target of the detection optical system are An example is coma. Since the interval between the mark images changes with high sensitivity to coma, coma can be corrected with high accuracy.

  Next, the adjustment method of the second position detection device described in the embodiment of the present invention includes an illumination system (1-8) for illuminating one or a plurality of test marks, and the test marks. And a detection optical system (10, 9, 12, 15, 16, 21) for condensing the light beam from the detection mark, and one of the test marks based on the light beam collected by the detection optical system , Or a position detecting device adjusting method for detecting the relative positions of the plurality of test marks, and a concave portion (33a, 35b) and a convex portion in a predetermined measuring direction on a predetermined substrate (11), respectively. (33b, 35a) are alternately and periodically arranged, and two lattice marks (HM1, HM2) having shapes in which the concave portions and the convex portions are inverted with respect to each other are formed in proximity to each other. Its measurement of those grid-like marks via its detection optics The spacing of the direction is measured, and adjusts the predetermined optical characteristics of the illumination system on the basis of the measured value.

  According to the present invention, if the width of the concave portion (33a) of the first lattice mark (HM1) of these lattice marks is narrower than the width of the convex portion (33b), for example, the concave portion (33a) is darkened. A level image is obtained. Correspondingly, in the second lattice mark (HM2), the width of the convex portion (35a) is narrower than the width of the concave portion (35b), so that an image having a dark level at the convex portion (35a) is obtained. That is, there is a difference in level between the first grid mark (HM1) and the second grid mark (HM2) where dark-level images can be obtained. For this reason, if there is an adjustment residual of the illumination system, for example, a position shift of the aperture stop or an uneven illuminance distribution at the position of the aperture stop, the measured value of the interval between the images of the two grid marks shifts. The adjustment residual of the illumination system can be corrected by adjusting the interval so as to drive it to a predetermined reference value. Also in this case, the lattice marks can be easily and accurately formed by using a predetermined mask.

In this case, an example of the optical characteristic to be adjusted in the illumination system is a position in a plane perpendicular to the optical axis of the aperture stop (3) in the illumination system.
Further, in the second adjustment method, the two lattice marks (HM1, HM2) on the substrate (11) are defined as the first lattice marks, and the concave portions are respectively formed in the measurement direction on the substrate. Convex portions are alternately and periodically arranged, and two second lattice marks (DM1, DM2) having different ratios between the widths of the concave portions and the convex portions are formed closer to each other. In addition, after adjusting the predetermined optical characteristic of the illumination system based on the interval between the first grid marks (HM1, HM2), the second grid pattern is passed through the detection optical system. An interval in the measurement direction of the marks (DM1, DM2) may be measured, and predetermined optical characteristics of the detection optical system may be adjusted based on the measurement values.

  This means that the second adjustment method of the present invention and the first adjustment method (different ratio mark method) are used in combination. At this time, the adjustment of the predetermined optical characteristic of the illumination system using the first lattice mark (HM1, HM2) is not affected by the aberration of the detection optical system. Therefore, the illumination system is first adjusted using the first lattice marks (HM1, HM2), and then, for example, the asymmetric aberration of the imaging optical system is adjusted by the different ratio mark method (first adjustment method). Therefore, it is convenient that both can be adjusted independently.

  In the first or second adjustment method, the plurality of lattice marks (DM1, DM2; HM1, HM2) are formed on the substrate in close proximity in the measurement direction. It is desirable. As a result, the measurement image interval between the mark images can be measured with high accuracy without so-called Abbe error, and the optical error can be corrected with high accuracy. Furthermore, it is desirable that the step of the concave portion (31a, 32a; 33a, 35b) and the step of the convex portion (31b, 32b; 33b, 35a) are substantially in the range of 40 to 60 nm, respectively. As a result, a high-contrast image can be obtained, so that the intervals between a plurality of marks can be detected with high accuracy.

  Next, the first position detection device described in the embodiment of the invention of the present application is a detection optical system (10, 9, 12,...) That collects light beams from one or a plurality of test marks. 15, 16, 21) and a photoelectric detector (22) for receiving the light beam condensed by the detection optical system, and the one test mark based on the detection signal of the photoelectric detector Or at least a portion of the optical member (16) that affects predetermined optical characteristics in the optical system for detection (for example, A positioning member (16a, 16b, 17a, 17b) for positioning in a plane perpendicular to the optical axis of the optical system for detection, a predetermined optical system that is detected via the optical system for detection, and the photoelectric detector. Measurement method for multiple grid marks Based on the interval for the control operation system for driving the positioning member in order to reduce the error in the predetermined optical characteristic (23), those having a. According to the present invention, the adjustment method of the first position detection device described in the embodiment of the present invention can be used.

  Next, the second position detection device described in the embodiment of the present invention includes an illumination system (1 to 8) for illuminating one or a plurality of test marks, and a light flux from the test marks. A detection optical system (10, 9, 12, 15, 16, 21) that collects light and a photoelectric detector (22) that receives the light beam collected by the detection optical system, In a position detection device that detects the position of one test mark or the relative position of the plurality of test marks based on the detection signal of the photoelectric detector, it affects the predetermined optical characteristics in the illumination system. A positioning member (4a, 4b, 5a, 5b) for positioning at least a part of the optical member (3) to be provided (for example, in a plane perpendicular to the optical axis of the illumination system), an optical system for detection thereof, and A plurality of predetermined grid-shaped detectors detected via a photoelectric detector Based on the interval for a given measurement direction of click, the control operation system for driving the positioning member so as to reduce the error in the predetermined optical characteristic (23), those having a. According to the present invention, the second position detecting apparatus adjustment method of the present invention can be used.

  Next, the adjustment method of the optical system described in the embodiment of the invention of the present application is based on the illumination system (1 to 3, 6 to 8) that irradiates the specimen with illumination light and the specimen. Is an optical system adjustment method for adjusting at least one predetermined optical characteristic with a detection optical system (10, 9, 12, 15, 16, 21) that collects the luminous flux of 11A) The first and second test marks (HM1, HM2; 28A) are formed on the predetermined positional relationship on the upper side, and the two test marks are rotated by a predetermined angle while maintaining the positional relationship. The third and fourth test marks (HM3, HM4; 28B) in the state are formed, and the first and second test marks (HM1, HM2) on the substrate through the detection optical system. Measuring the relative position of 28A) and detecting the light without rotating the substrate; The relative positions of the third and fourth test marks (HM3, HM4; 28B) on the substrate are measured through the system, and the illumination is based on the relative positions measured for the two sets of test marks. The system or at least one of the detection optical system is adjusted.

  According to the present invention, first, after measuring the relative positions (for example, the interval D1) of the first and second test marks on the substrate, the substrate is rotated without rotating the substrate. The relative positions of the third and fourth test marks (for example, the interval D2) are measured. As a result, TIS (Tool Induced Shift), which is an error caused by the apparatus, is (D1-D2) / 2 as an example, and the illumination system or the detection optical system is adjusted so that the error falls within a predetermined allowable range. At least one adjustment is made. At this time, it is not necessary to rotate the substrate, and a normal stage that can be positioned in two dimensions is used as long as the two sets of test marks are sequentially stored in the observation field of the detection optical system. Therefore, the error can be measured with high accuracy in a short time.

In this case, as an example, the illumination system, or the above-mentioned so that the interval measured for the first and second test marks is equal to the interval measured for the third and fourth test marks. Adjustment of at least one of the detection optical systems may be performed. As a result, adjustment is performed so that the TIS is substantially minimized.
A pair of box-in-box marks (28A) may be used as the first and second test marks. In this case, for example, the TIS of the overlay error measuring device is measured.

  Further, as the first and second test marks, the concave portions (33a, 35b) and the convex portions (33b, 35a) are alternately and periodically arranged in a predetermined direction, and the concave portions and the convex portions thereof are mutually arranged. A pair of lattice marks (HM1, HM2) having a shape inverted from the portion may be used. In this case, for example, an error caused by a positional deviation of the center of the aperture stop of the illumination system can be measured and adjusted with high accuracy.

  In addition, the first and second test marks are formed by arranging the concave portions (31a, 32a) and the convex portions (31b, 32b) at a predetermined pitch in a predetermined direction, respectively. Even if a pair of grid marks (DM1, DM2) having different duty ratios expressed as a percentage of the ratio between the width and the width of the convex portion, and thus the ratio of the width of the concave portion (or the convex portion) to one pitch, is used. Good. In this case, an error caused by asymmetric aberration such as coma aberration of the detection optical system can be measured and adjusted.

Further, the third and fourth test marks on the substrate are desirably marks obtained by rotating the first and second test marks by 180 °. Thereby, the TIS as defined in the past can be measured.
The first evaluation substrate described in the embodiment of the invention of the present application is an evaluation substrate (11A; 11B) on which a plurality of test marks are formed. The test marks (HM1, HM2, 28A; DM1, DM2) are formed in a predetermined positional relationship, and the two test marks are rotated in a predetermined angle while maintaining the positional relationship. A fourth test mark (HM3, HM4, 28B; DM2, DM4) is formed. By using this substrate, the optical system adjustment method described in the embodiment of the present invention can be implemented.

  Next, the second evaluation substrate described in the embodiment of the present invention is an evaluation substrate (11) on which a plurality of test marks are formed, and the concave portion and the convex portion are provided. At least two first test marks (DM1, DM2) that are alternately arranged and have different ratios between the widths of the concave portions and the convex portions are formed. When this substrate is used, the position detection device can be adjusted by the above-described different ratio mark method.

  In addition, the substrate is used, for example, for adjusting an optical apparatus incorporated in an apparatus used in a device manufacturing process including a lithography process in which a device pattern is transferred directly or onto a workpiece (W) via a mask (R). It is done. As an example, the substrate has substantially the same shape and size as an object to be detected by the optical device, and thus it is not necessary to newly manufacture a holder or the like.

  Moreover, the pattern detection apparatus described in the embodiment of the invention of the present application includes an illumination system (1 to 3, 6 to 8) that irradiates a test object with illumination light via an objective optical system (10), and In a pattern detection apparatus having a detection system (9, 12, 15, 16, 18, 21) that receives a light beam generated from a test object and passing through the objective optical system (10), a pair arranged in the first direction. First marks (HM1, HM2) and a second pair of second lines arranged in the second direction intersecting with the first direction (including the case of 180 ° rotation) and having the same configuration as the pair of first marks. A movable member (11A) integrally provided with marks (HM3, HM4; 25Y), relative position information obtained by detecting the pair of first marks via the objective optical system, and the pair of pairs Based on the relative position information obtained by detecting the second mark An adjustment mechanism (4a, 4b, 5a, 5b, 16a, 16b, 17a, 17b) for adjusting the illumination system, the objective optical system, and at least a part of the optical system (3, 16) in the detection system; It is provided.

According to such a pattern detection apparatus, the optical system adjustment method described in the embodiment of the present invention can be used. Furthermore, the optical characteristics can be adjusted two-dimensionally.
Next, the adjustment method of the third position detection device described in the embodiment of the present invention includes an illumination system (1 to 3, 6 to 8) for illuminating one or a plurality of test marks, And a detection optical system (10, 9, 12, 15, 16, 21) for condensing the light beam from the test mark, one of which is based on the light beam collected by the detection optical system. An adjustment method of a position detection device for detecting the position of one test mark or the relative position of a plurality of test marks, comprising a center portion (DM22; HM22) made of a concavo-convex pattern and a predetermined center portion. Substrate (11C) on which evaluation marks (DX, HX) having two end portions (DM21, DM23; HM21, HM23) each having an uneven pattern arranged symmetrically so as to be sandwiched along the measurement direction are formed Optical system test for detecting The relative positional relationship (interval, deviation, etc.) of the central portion and the two ends with respect to the measurement direction is detected via the detection optical system, and the detection result is Based on this, the predetermined optical characteristics of the illumination system or the detection optical system are adjusted.

  According to the present invention, as an example, the intervals between the center portion of the evaluation mark and the end portions on both sides are detected, and these intervals are respectively compared with a predetermined reference value (substantially true value). Thus, an error (TIS) caused by the apparatus is obtained. At this time, as a method for determining the reference value, a method is conceivable in which the evaluation mark is rotated by 180 °, the interval between them is measured again, and the average value of the two measurement results is used as the reference value. . In this case, the evaluation mark described in the embodiment of the invention of the present application is symmetric (line symmetric) in the measurement direction with respect to the center portion, and therefore, even if the evaluation mark is rotated 180 °, the measurement direction The shapes of are substantially the same. Therefore, errors such as distortion other than the error caused by the apparatus are not mixed, and only the error caused by the apparatus can be obtained with high accuracy, and the error can be adjusted with high accuracy.

Further, when the central part and the two end parts constituting the evaluation mark are lattice marks (different ratio marks) in which the ratio of the width of the concave part to the width of the convex part is different from each other, In addition to being able to form accurately, as an example, the imaging optical state of the optical system for detection, particularly the coma error, can be measured with high accuracy.
On the other hand, when the central portion and the two end portions constituting the evaluation mark are lattice marks (different step marks) in which the concave portion and the convex portion are reversed, the illumination state of the illumination system is Deviations (illumination aperture stop deviation, illuminance distribution non-uniformity, etc.) can be measured with high accuracy.

  Next, the third position detection device described in the embodiment of the present invention includes an illumination system (1-3, 6-8) for illuminating one or a plurality of test marks, and the test thereof. A detection optical system (10, 9, 12, 15, 16, 21) for condensing the light beam from the mark, and a photoelectric detector (22) for receiving the light beam collected by the detection optical system; In a position detection device that detects the position of the one test mark or the relative positions of the plurality of test marks based on the detection signal of the photoelectric detector, the illumination system and the detection system A positioning member that positions at least some of the optical members (3; 16) that affect predetermined optical characteristics in the optical system, a detection optical system, and a predetermined detection that is detected via the photoelectric detector At least three of the evaluation marks (DM21 to DM 3; HM21~HM23) based on the relative positional relationship, a control operation system for driving the positioning member in order to reduce the error in the predetermined optical characteristic (23), those having a. With this apparatus, the third position detecting apparatus adjustment method of the present invention can be used.

  Next, the exposure apparatus of the present invention includes the above-described position detection apparatus of the present invention, a mask stage (54, 55) for holding a mask, and a mask pattern is transferred and an alignment mark for alignment is formed. And a substrate stage (58, 59) for positioning the substrate (W), and the position detection device detects the position of the alignment mark on the substrate, and based on the detection result. The mask and the substrate are aligned. By adjusting the optical system of the position detection apparatus of the present invention using the adjustment method of the present invention, high overlay accuracy can be obtained.

  The device manufacturing method of the present invention is a device manufacturing method for manufacturing a predetermined device using the position detecting device adjusting method of the present invention, and the predetermined position detecting device using the adjusting method. The position of the alignment mark on the predetermined substrate is detected using the position detection device after the adjustment, and the substrate and the mask are aligned based on the detection result. The method includes a step of transferring the mask pattern onto the substrate. In this case, since high overlay accuracy can be obtained, high-functional devices can be mass-produced with a high yield.

  Next, an exposure apparatus manufacturing method described in the embodiment of the present invention is an exposure apparatus manufacturing method for exposing a photosensitive substrate (W) with an energy beam through a mask (R). Is provided with a mark detection system (63) for detecting alignment marks (38, 40X, 40Y) on the substrate so that the optical axis is arranged outside the irradiation region of the energy beam on the coordinate system in which In order to detect an interval in the arrangement direction of at least two test marks in which concave portions and convex portions are alternately arranged, the mark detection system detects the at least two test marks, and the mark detection system In order to adjust the optical characteristics, at least one optical element of the mark detection system is moved or replaced based on the detected interval.

In this case, since the mark detection system can be adjusted with high accuracy by applying an adjustment method such as the different ratio mark method described in the embodiment of the present invention, high overlay accuracy can be obtained.
Next, the adjustment method described in the embodiment of the invention of the present application includes an irradiation system that irradiates the object with illumination light, and a detection optical system that condenses the light beam from the object. A method for adjusting an optical system in a position detection apparatus, wherein the first optical characteristic of the irradiation system is adjusted, and after the first optical characteristic is adjusted, the second optical characteristic of the detection optical system is adjusted. Is to do.

  The exposure method described in the embodiment of the invention of the present application is an exposure method in an exposure apparatus including a position detection device that is an object of the adjustment method, and the position detection device adjusted by the adjustment method is used. The alignment mark formed on the substrate is detected, the substrate is aligned based on the mark detection result, and a predetermined pattern is exposed on the aligned substrate.

  Further, another position detection device described in the embodiment of the present invention includes an irradiation system that irradiates an object with illumination light, and a first adjustment that adjusts a first optical characteristic of the irradiation system. Apparatus, detection optical system for condensing a light beam from the test object, a second adjustment device for adjusting a second optical characteristic of the detection optical system, and adjustment of the first optical characteristic by the first adjustment device And a control device for adjusting the second optical characteristic by the second adjusting device later.

  Further, another exposure apparatus described in the embodiment of the present invention is a substrate using the further position detection apparatus of the present invention and the position detection apparatus adjusted by the first and second adjustment apparatuses. An alignment device that detects an alignment mark formed on the alignment mark and aligns the substrate based on the detection result of the alignment mark, and exposes a predetermined pattern on the aligned substrate. It is.

According to the measurement method of the present invention, a test mark for characteristic measurement can be easily and accurately formed.
According to the adjustment method of the present invention, it is possible to accurately form a mark for characteristic measurement, and to accurately correct, for example, a predetermined aberration of a detection optical system or an adjustment residual of an illumination system. As a result, the error (TIS) caused by the apparatus can be adjusted with high accuracy.

Hereinafter, a preferred first embodiment of the present invention will be described with reference to FIGS.
FIG. 1 shows a position detection apparatus of this example. In FIG. 1, the surface of an adjustment wafer 11 is arranged on a surface to be measured by the position detection apparatus of this example. A plurality of pairs of concave and convex lattice marks (step marks) are formed on the surface of the wafer 11 as described later.
In FIG. 1, broadband illumination light AL1 emitted from a light source 1 such as a halogen lamp, along a light axis AX1 of the illumination system, is a condenser lens 2 and an aperture stop (hereinafter referred to as “σ stop”) 3 of the illumination system. The light enters the half prism 9 through the first relay lens 6, the field stop 7 and the second relay lens 8. The illumination light AL 1 reflected downward by the half prism 9 illuminates the adjustment wafer 11 through the objective lens group 10. The σ stop 3 sets a σ value that is a coherence factor of the illumination system.

  The light beam AL2 reflected by the wafer 11 passes through the objective lens group 10 and enters the half prism 9. Then, the light beam AL2 transmitted through the half prism 9 forms an image of a lattice mark on the wafer 11 on the pattern surface of the index plate 13 via the third relay lens 12. That is, the pattern surface of the index plate 13 is conjugate with the surface of the wafer 11, and index marks 14a and 14b serving as reference positions for detection positions when alignment is performed in the projection exposure apparatus are formed on the pattern surface. However, when the position detection device of this example is used as an overlay error measurement device, the index plate 13 is not necessarily required.

  The light beam AL2 that has passed through the index plate 13 passes through the fourth relay lens 15, the coma aberration correcting optical system (hereinafter referred to as “coma correcting optical system”) 16, the aperture stop 18, the field lens 21, and the CCD type or the like. On the imaging surface of the two-dimensional imaging element 22, images of lattice marks and index marks 14 a and 14 b on the wafer 11 are formed. The image signal S read from the image sensor 22 is supplied to the control calculation system 23. In this example, an imaging optical system as a detection optical system is configured by the objective lens group 10, the half prism 9, the third relay lens 12, the fourth relay lens 15, the coma correction optical system 16, and the field lens 21. Has been. Hereinafter, taking the Z axis parallel to the optical axis AX2 of this imaging optical system, taking the X axis parallel to the paper surface of FIG. 1 in the plane perpendicular to the Z axis, and the Y axis perpendicular to the paper surface of FIG. explain.

  In this case, the aperture stop 18 in the imaging optical system determines the numerical aperture (NA) of the imaging optical system according to the aperture diameter, and the wafer 11 with respect to the best focus position of the imaging optical system according to the position. This affects the deviation of the detection result of the mark position (hereinafter referred to as “telecentric deviation”) accompanying the positional deviation amount (defocus amount) in the direction of the optical axis AX2 of the surface. Therefore, the aperture stop 18 is held by the holding members 19a and 19b, and under the control of the control calculation system 23, a feed screw method, a telescopic drive element method such as a piezo element, a voice coil motor (VCM) method. Alternatively, the center of the aperture stop 18 can be shifted by an arbitrary amount in the X and Y directions by shifting the holding members 19a and 19b by the aperture stop position adjusting mechanisms 20a and 20b such as a linear motor system. Yes. Before the measurement is started, the position of the aperture stop 18 is adjusted in advance so that the telecentric shift does not occur.

  In the position detection apparatus of this example, the objective lens group 10 and other optical systems are all manufactured and assembled with extremely high accuracy, but various aberrations due to manufacturing errors still remain. If such residual aberration exceeds a predetermined allowable range, it will deteriorate the position detection accuracy beyond the required accuracy. The aberration that deteriorates the position detection accuracy is mainly coma aberration, which occurs due to decentering of each optical member and insufficient accuracy of the polished surface. Therefore, even when the test object (test mark) is on the optical axis AX2 of the imaging optical system, the coma aberration is adversely affected.

  The coma correction optical system 16 is an optical system for correcting such residual coma aberration, and the coma aberration state of the imaging optical system can be changed by changing the position thereof. Therefore, the frame correction optical system 16 is held by the holding members 16a and 16b, and the holding members 16a and 16b are shifted by the frame correction optical system position adjusting mechanisms 17a and 17b under the control of the control calculation system 23. The center of the frame correction optical system 16 can be shifted by an arbitrary amount in the X direction and the Y direction. In addition, by moving at least one optical element other than the coma correction optical system 16, the optical characteristics of the imaging optical system, that is, the aberration other than the coma aberration, the telecentricity, the focal position, and the like can be adjusted. The at least one optical element may be moved so as to substantially cancel the fluctuation of the optical characteristics that may be caused by the movement of the correction optical system 16.

By the way, when the coma remaining in the optical system is very large, the position adjustment amount of the coma correction optical system 16 increases, and the position of the principal ray of the imaging light beam may be greatly shifted from the optical axis AX2. is there.
The positional relationship between the position of the aperture stop 18 and the principal ray changes greatly. In the worst case, it is necessary to readjust the position of the aperture stop 18 every time the position of the coma correction optical system 16 is adjusted. May occur.

Therefore, the positional relationship between the aperture stop 18 and the frame correction optical system 16 is reversed from that in FIG. 1, and the aperture stop 18 is disposed closer to the wafer than the frame correction optical system 16, so that You can eliminate any concerns about adjustment.
The frame correction optical system 16 is preferably arranged in the vicinity of the pupil plane (aperture stop 18) in the imaging optical system, but may of course be arranged on the conjugate plane via a relay system. By arranging on the conjugate plane via the relay optical system, there is an advantage that the degree of freedom with respect to the space to be arranged is increased.

  On the other hand, when the relationship between the position of the σ stop 3 in the illumination system from the light source 1 to the second relay lens 8 and the position of the aperture stop 18 in the imaging optical system changes, the detection position of the test mark is detected. Adversely affected. However, as described above, when the position of the aperture stop 18 is adjusted (moved), a telecentric displacement occurs. Therefore, the positional relationship between them is adjusted by adjusting the position of the σ stop 3. For this purpose, the σ diaphragm 3 is held by the holding members 4a and 4b, and the holding members 4a and 4b are shifted by the σ diaphragm position adjusting mechanisms 5a and 5b under the control of the control calculation system 23, thereby obtaining the σ diaphragm. 3 can be shifted by an arbitrary amount in directions corresponding to the X direction and the Y direction on the wafer 11.

Next, an example of a method for adjusting the optical system of the position detection apparatus in FIG. 1 will be described. First, a plurality of pairs of concave and convex lattice marks (step marks) on the adjustment wafer 11 of this example will be described.
FIG. 2 is a plan view showing the wafer 11 for adjustment in FIG. 1. In FIG. 2, a silicon wafer is used as the wafer 11 as an example. On the surface of the wafer 11, a pair of X-axis marks (DM1, DM2) each composed of a first mark DM1 and a second mark DM2, each of which is composed of a periodic uneven pattern in the X direction and arranged in series, In addition, a pair of Y-axis marks (DM11, DM12) including a first mark DM11 and a second mark DM12, which are formed by rotating the pair of marks by 90 °, are formed. These two pairs of marks (DM1, DM2) and marks (DM11, DM12) are used for adjusting the characteristics of the optical system for detection.

  Also, on the surface of the wafer 11, a first mark HM1 composed of a periodic uneven pattern in the X direction, and a second mark which is arranged in series in the X direction with respect to this mark and has an inverted uneven pattern. A pair of X-axis marks (HM1, HM2) formed with the mark HM2, and a pair of Y-axis marks (HM11, HM12) formed by rotating the pair of marks by 90 °, a first mark HM11, and a second mark HM12 HM12) is also formed. These two pairs of marks (HM1, HM2) and marks (HM11, HM12) are used for adjusting the characteristics of the illumination system.

For example, when a silicon wafer is used as the wafer 11, these uneven lattice marks are formed by applying a photoresist onto the surface, exposing a projected image of a corresponding reticle pattern, developing a photoresist, etching, and resist. It can be formed with extremely high accuracy by a process such as peeling.
3A is an enlarged plan view showing a pair of marks (DM1, DM2) on the X axis for adjusting the characteristics of the optical system for detection shown in FIG. 2, and FIG. 3B is a diagram of FIG. The first mark DM1 is a cross-sectional view of a pattern in which five long and thin linear concave patterns 31a having a line width a are formed on the surface of the wafer 11 in a grid pattern with a predetermined step Hd and a pitch P in the X direction. The pitch P is about 5 to 20 μm. Similarly, the second mark DM2 is a pattern in which five linear concave patterns 32a having a line width c are formed on the surface of the wafer 11 in a grid pattern at the same pitch P in the X direction at the level difference Hd. is there. In the X direction that is the measurement direction, the distance between the center of the first mark DM1 and the center of the second mark DM2 is set to Dd as a design value. This distance Dd is about 50 to 100 μm.

  The step (depth) Hd of each of the marks DM1 and DM2 is preferably about 40 to 100 nm. If the level difference Hd is too small, the contrast of those mark images is lowered (the mark portion is not sufficiently dark), and the position detection accuracy is lowered. On the other hand, when the step Hd is higher than about 100 nm, adverse effects such as geometric optical vignetting due to the step portion occur, and high-precision measurement becomes difficult. Further, in order to obtain an image with good contrast for each mark, the step Hd is desirably 40 to 60 nm.

  In this example, the ratio (a: b or a / b) between the width a of the concave pattern 31a of the first mark DM1 and the width b (a + b = P) of the convex pattern 31b, and the concave pattern of the second mark DM2. The ratio (c: d or c / d) of the width c of 32a and the width d (c + d = P) of the convex pattern 32b is different. In other words, the duty ratio (100 × a / P (%)) of the width a of the concave pattern 31a of the first mark DM1 to the pitch P and the duty ratio of the width c of the concave pattern 32a of the second mark DM2 to the pitch P ( 100 × c / P (%)). As an example, in this example, the duty ratio of the concave pattern 31a of the first mark DM1 is set to 50%, and the ratio between the width a of the concave pattern 31a and the width b of the convex pattern 31b is set as follows. Yes.

a: b = 1: 1 or a / b = 1 (11)
On the other hand, the duty ratio of the concave pattern 32a of the second mark DM2 is set to about 10%, and the ratio between the width c of the concave pattern 32a and the width d of the convex pattern 32b is set as follows.
c: d = 1: 9, or c / d = 1/9 (12)
When the two marks DM1 and DM2 arranged close to each other in the X direction are observed with the position detection device of FIG. 1, image signals obtained by reading out the images in the X direction by the image sensor 22 are obtained. The image signal SD shown in FIG. The image signal SD may be an image signal obtained by scanning the two mark images in FIG. 3A in the X direction and averaging in the non-measurement direction (Y direction).

  In FIG. 3C, the horizontal axis represents the position X in the measurement direction, and the position X is actually imaged using the magnification α from the surface of the wafer 11 to the imaging surface of the imaging element 22 in FIG. This represents a position obtained by multiplying the position from a predetermined reference point on the imaging surface of the element 22 by 1 / α. In the image signal SD of FIG. 3C, the edge portion of the first mark DM1 becomes a dark portion in the portion ID1 corresponding to the image of the first mark DM1, and the same applies to the portion ID2 corresponding to the image of the second mark DM2. The edge part is a dark part.

  The control calculation system 23 in FIG. 1 obtains an interval Md between the center position Xd1 of the part ID1 and the center position Xd2 of the part ID2 from the image signal SD. This distance Md should be the designed distance Dd in FIG. 3A when there is no aberration in the detection optical system. However, if coma aberration remains in the detection optical system, the amount of change in image position due to coma aberration varies depending on the duty ratio of the widths of the concave patterns 31a and 32a of both marks DM1 and DM2. Md does not match Dd. Further, not only the magnitude of the residual coma aberration but also the sign can be determined from the magnitude relationship between the measured interval Md and the reference interval Dd. Therefore, an error ΔMd (= Dd−) with respect to the interval Dd of the interval Md. Based on the sign and magnitude of Md), the control calculation system 23 in FIG. 1 adjusts the position of the frame correction optical system 16 via the frame correction optical system position adjustment mechanisms 17a and 17b so that the error ΔMd becomes small. To do.

  After adjusting the position of the frame correction optical system 16, the distance Md between the images of the marks DM1 and DM2 is measured again. If the error ΔMd with respect to the reference value Dd is out of the allowable range, the above adjustment is performed again. Do. By repeating the above steps until the error ΔMd falls within the allowable range, the adjustment of the coma aberration of the detection optical system is completed. The adjustment method of this example is a method of measuring mark intervals with different duty ratios, and is therefore referred to as “different ratio mark method”.

  Note that there are various methods such as a slice method and a correlation method in the position detection algorithm of each mark image in the above-described position detection process, and any of them may be used in this example. For example, in the slice method, position detection is performed based on the slice position of the image signal SD at a predetermined slice level, and in the correlation method, the image signal SD is compared with a predetermined reference waveform, and the degree of correlation with the reference waveform is determined. The highest position is the mark image position. In FIG. 3C, signal waveforms corresponding to the images of the index marks 14a and 14b are not shown. However, for example, after detecting the positional deviation amounts of the marks DM1 and DM2 with respect to the index marks 14a and 14b, The interval Md may be obtained.

  In addition, the distance Dd between the marks DM1 and DM2 in FIG. 3A is known as a design value, but the actual mark distance Dd is the pattern of the reticle used to transfer these marks. Manufacturing errors and manufacturing errors such as an etching error when forming a step on the wafer 11 are mixed. Therefore, it is desirable to measure the actual value of the reference interval Dd before the adjustment. For this purpose, for example, T. Kanda, K. Mishima, E. Murakami and H. Ina: Proc. SPIE, Vol. 3051, pp. 846-855 (1997) (hereinafter referred to as “Document 1”) cited in the prior art. 3), the distance between the images of the two marks DM1 and DM2 (referred to as Md1) in FIG. 3A is measured, and then the wafer 11 is rotated by 180 ° to obtain the same two marks DM1. , DM2 image interval (Md2) is measured, and an average value <Md> of these two measurement values may be adopted instead of the reference interval Dd, as represented by the following equation.

<Md> = (Md1 + Md2) / 2 (13)
Thereby, even if there is residual aberration in the detection optical system of FIG. 1, the error ΔMd of the interval Md between the two mark images due to coma aberration can be detected with high accuracy.
By the way, in general, a position detection device needs to measure a mark position in a two-dimensional direction (X direction, Y direction) or a relative positional relationship. Therefore, similarly to the adjustment in the X direction, the coma aberration in the Y direction is also adjusted by measuring the distance between the images of the pair of marks (DM11, DM12) on the Y axis on the wafer 11 in FIG. Can do.

Next, a method for adjusting the illumination system in FIG. 1 will be described. For this purpose, first, a pair of X-axis marks (HM1, HM2) on the wafer 11 for adjustment shown in FIG. 2 is used.
4A is an enlarged plan view showing a pair of X-axis marks (HM1, HM2) for adjusting the characteristics of the illumination system shown in FIG. 2, and FIG. 4B is a sectional view of FIG. 4A. Yes, the first mark HM1 is a pattern in which five linear concave patterns 33a having a line width e are formed on the surface of the wafer 11 in a grid pattern with a predetermined step Hh and a pitch P2 in the X direction. The pitch P2 is about 5 to 20 μm. On the other hand, the second mark HM2 is a pattern in which the outer periphery thereof is surrounded by the engraved portion 34, and five elongated linear convex patterns 35a having a line width e are formed in a lattice shape at a pitch P2.

  And the design value of the space | interval to the X direction of the center of both marks HM1 and HM2 is Dh, and the space | interval Dh is about 50-100 micrometers. In this case, the first mark HM1 is a mark in which the width of the concave pattern 33a is e and the width of the convex pattern 33b is f (= P2-e), and the second mark HM2 is the width of the convex pattern 35a is e. The concave pattern 35b is a mark having a width f, and both marks HM1 and HM2 are marks in which only the concavo-convex relationship is reversed. That is, the duty ratio of the concave pattern 33a of the first mark HM1 and the duty ratio of the convex pattern 35a of the second mark HM2 are set to be equal.

  When the two marks HM1 and HM2 arranged close to each other as described above are observed by the position detection device of FIG. 1, image signals obtained by reading out these images in the X direction by the image sensor 22 (or The signal averaged in the non-measurement direction) is defined as an image signal SH in FIG. In the image signal SH of FIG. 4C, in the portions IH1 and IH2 corresponding to the first mark HM1 and the second mark HM2, the mark portions are dark portions. 1 detects the position Xh1 of the center of the portion IH1 and the position Xh2 of the center of the portion IH2 of the image signal SH, and obtains the interval Mh. Also in this case, the interval Mh is a value obtained by multiplying the interval on the imaging surface by 1 / α using the magnification α from the surface of the wafer 11 to the imaging surface in FIG.

In a state where there is no adjustment residual in the illumination system, the interval Mh thus determined should be equal to the reference interval Dh, but if there is an adjustment error in the illumination system, the two observed Since the shift amount of the image position of the marks HM1 and HM2 differs depending on the step of each mark, the measured interval Mh is different from the reference interval Dh.
Specifically, in this example, the sign of the error ΔMh (= Dh−Mh) with respect to the reference interval Dh of the image interval Mh of the two marks HM1 and HM2 in accordance with the position of the σ stop 3 of the illumination system in FIG. And the size changes. Therefore, the control calculation system 23 in FIG. 1 adjusts the position of the σ stop 3 via the σ stop position adjusting mechanisms 5a and 5b so that the error ΔMh becomes small. Thereafter, the error ΔMh of the image interval Mh between the two marks HM1 and HM2 is measured again, and the position of the σ stop 3 is adjusted by adjusting the position of the σ stop 3 until the error ΔMh falls within the allowable range. Complete.

The step Hh of the marks MH1 and MH2 is desirably about 40 to 100 nm, like the step Hd of the marks MD1 and MD2 in FIG. The reason is the same as above. Furthermore, in order to obtain an image signal SH with better contrast, the step Hh is also preferably about 40 to 60 nm.
The duty ratio (100 × e / P2 (%)) of the concave pattern 33a of the first mark HM1 and the duty ratio (100 × e / P2 (%)) of the convex pattern 35a of the second mark HM2 are 10%. It is desirable that the degree. This is also because if the duty ratio is too small, the contrast of the mark image is lowered and the reproducibility of the position detection result is deteriorated. If the duty ratio is too large, relative to the image of the concave pattern 33a of the first mark HM1 and the image of the convex pattern 35a of the second mark HM2 due to the displacement of the σ stop 3 (due to the adjustment residual of the illumination system). This is because the amount of change in the amount of misalignment becomes small, and the adjustment sensitivity decreases.

Similarly to the adjustment in the X direction described above, the adjustment residual of the illumination system in the Y direction is measured by measuring the distance between the images of the pair of Y-axis marks (HM11, HM12) on the wafer 11 in FIG. Adjustments can also be made.
Further, in the case of FIG. 4 as well, it is desirable to measure the actual value of the interval Dh serving as the reference. For this purpose, as in the case of FIG. 3, after measuring the distance between the images of the two marks HM1 and HM2 (referred to as Mh1) in FIG. The interval between the images of the marks HM1 and HM2 (referred to as Mh2) may be measured, and the average value of these two measured values may be adopted instead of the reference interval Dh.

  In the above embodiment, the position adjustment of the coma aberration correcting optical system 16 and the position adjustment of the σ stop 3 in FIG. 1 may be performed independently. However, when the position of the σ stop 3 is adjusted, even if coma remains in the detection optical system (imaging optical system), the adjustment can be performed without being affected by the coma. It is efficient to adjust the coma aberration by moving the coma correction optical system 16 after adjusting the position of the σ stop 3 first.

In this example, the coma correction optical system 16 in FIG. 1 is adjusted to adjust and remove the remaining coma aberration. However, the present invention is not limited to this, and other optical elements such as the objective lens group 10 and the half prism 9 are used. The remaining coma aberration may be adjusted and removed by adjusting the position or rotation angle of the member. Further, when adjusting the illumination state, not only the position of the σ stop 3 but also the position of the light source 1 or the position or rotation angle of the first relay lens 6 or the second relay lens 8 may be adjusted. Good.

[Second Embodiment]
Next, a second embodiment of the present invention will be described with reference to FIGS. 1 and 5 to 10. Since the position detection apparatus of FIG. 1 is also the object of adjustment in this example, description of the apparatus configuration is omitted, and an adjustment method of the optical system will be described. In this example, a wafer 11A is installed instead of the adjustment wafer 11 shown in FIG. A plurality of pairs of test marks are formed on the surface of the wafer 11A. First, the plurality of pairs of test marks will be described.

  FIG. 5 is a plan view showing an adjustment wafer 11A installed in place of the wafer 11 in FIG. 1. In FIG. 5, a silicon wafer is used as an example of the wafer 11A. On the surface of the wafer 11A, a first box-in-box mark 28A is formed by arranging an inner frame mark 27A and an outer frame mark 27B surrounding the inner frame mark 27A in a predetermined positional relationship. Then, in the vicinity of the first box-in-box mark 28A, the pair of marks 27A and 27B are integrally rotated by 180 ° while maintaining the positional relationship between the inner frame mark 27A and the outer frame mark 27B. A second box-in-box mark 28B made up of the inner frame mark 27C and the outer frame mark 27D in a rotated arrangement is formed.

  FIG. 6 is an enlarged plan view showing the box-in-box marks 28A and 28B in FIG. 5. In FIG. 5, the first box-in-box mark 28A is a rectangular frame-like convex pattern, respectively. It is composed of an inner frame mark 27A and an outer frame mark 27B. Similarly, the second box-in-box mark 28B includes an inner frame mark 27C and an outer frame mark 27D having the same shape as the inner frame mark 27A and the outer frame mark 27B, respectively. In this case, in the first box-in-box mark 28A, the positional deviation amount (design value) in the X direction and Y direction of the center of the inner frame mark 27A with respect to the center A of the outer frame mark 27B is (dX1, dY1), in the second box-in-box mark 28B, the positional deviation amount (design value) in the X direction and Y direction of the center of the inner frame mark 27C with respect to the center B of the outer frame mark 27D is (dX2, If dY2), the following relationship is established in terms of design.

(DX2, dY2) =-(dX1, dY1) (21)
In practice, there are mask pattern drawing errors and manufacturing process errors, but such errors may be smaller than the detection accuracy required by the position detection apparatus. These misregistration amounts may be measured and confirmed in advance with a scanning electron microscope (SEM) or the like, for example.

In this case, if the amount of positional deviation of the inner frame mark relative to the outer frame mark of the box-in-box marks 28A, 28B is detected via the position detection device of FIG. The relationship of equation (21) may not be established for the measured positional deviation amount. As will be described later, the asymmetrical aberration and the like can be adjusted using this.
Further, in FIG. 5, on the surface of the wafer 11A, a first mark HM1 made of a periodic uneven pattern in the X direction, and an uneven pattern arranged in series in the X direction with respect to this mark are arranged. A pair of X-axis marks 25X composed of the inverted second marks HM2 and a pair of marks 26X arranged close to the pair of marks 25X and rotated by 180 ° from the pair of marks 25X. Is formed. That is, the pair of marks 26X includes the third mark HM3 and the fourth mark HM4 that are obtained by rotating the first mark HM1 and the second mark HM2 together by 180 ° while maintaining the mutual positional relationship. Yes. If the design distance in the X direction between the first mark HM1 and the second mark HM2 is Dh, the design distance in the X direction between the third mark HM3 and the fourth mark HM4 is also Dh.

  In FIG. 5, the pair of marks 26 </ b> X are arranged close to the pair of marks 25 </ b> X in the measurement direction (X direction), but in the non-measurement direction (Y direction) with respect to the pair of marks 25 </ b> X. You may arrange | position close. Further, on the wafer 11A, a pair of marks 25X and a pair of marks 26X are further integrally formed, and a pair of Y-axis marks 25Y and a pair of marks 26Y are formed by rotating 90 °. These four pairs of marks 25X, 26X, 25Y and 26Y are used for adjusting the characteristics of the illumination system as an example.

For example, when a silicon wafer is used as the wafer 11A, the above-described uneven box-in-box mark and grid-like mark are formed by applying a photoresist onto the surface, exposing a projection image of a corresponding reticle pattern, and a photoresist. The film can be formed with extremely high accuracy by processes such as development, etching, and resist stripping.
Next, an example of a method for adjusting the detection optical system (imaging optical system) of the position detection apparatus in FIG. 1 will be described. For this purpose, the outer frames of the two box-in-box marks 28A and 28B on the wafer 11A in FIG. 5 are placed in the vicinity of the center (optical axis AX2) of the field of view of the objective lens group 10 of the position detection apparatus in FIG. The center position of the inner frame mark 27A is shifted relative to the center of the outer frame mark 27B, and the inner frame mark 27C is moved relative to the center of the outer frame mark 27D. The amount of positional deviation (δX2, δY2) at the center of is measured. This measured value is the reciprocal of the magnification α from the surface of the wafer 11A to the imaging surface of the image sensor 22 (the reciprocal of the position on the image sensor 22 obtained by processing the image signal S of the image sensor 22 in FIG. 1 / α). Further, the wafer 11A can be moved in the X direction or the Y direction at high speed using a wafer stage provided in the projection exposure apparatus or an XY stage provided in the overlay error measurement apparatus. .

Thereafter, the control calculation system 23 in FIG. 1 calculates the deviation amounts (δX, δY) from the equation (21) of the measured values of the positional deviation amounts of the two box-in-box marks 28A, 28B as follows. calculate. This deviation amount (δX, δY) corresponds to a part of TIS (Tool Induced Shift) of the detection optical system of the position detection device of FIG.
δX = (δX1 + δX2) / 2 (22A)
δY = (δY1 + δY2) / 2 (22B)
Then, the control calculation system 23 uses the coma correction optical system position adjustment mechanisms 17a and 17b as an example so that the deviation amount (δX, δY) approaches (0, 0) in the X direction. Adjust the position in the Y direction.

  Thereafter, the centers of the outer box marks of the two box-in-box marks 28A and 28B on the wafer 11A of FIG. 5 are sequentially moved again within the observation field of view of the position detection apparatus of FIG. The center position shift amount (δX1, δY1) of the mark 27A and the center position shift amount (δX2, δY2) of the inner frame mark 27C are measured. Then, the position adjustment of the coma correction optical system 16 and the position shift amount are calculated until the shift amounts (δX, δY) obtained by substituting these values into the equations (22A) and (22B) are within a predetermined allowable range. Repeat measurement. This completes the adjustment of coma, which is the main part of the asymmetric aberration of the detection optical system.

Next, a method for adjusting the illumination system in FIG. 1 will be described. For this purpose, first, two pairs of marks 25X and 26X on the X axis on the adjustment wafer 11A of FIG. 5 are used. Then, the center of one pair of marks 25X is moved to the vicinity of the center of the observation visual field of the position detection device of FIG.
7A is an enlarged plan view showing one pair of marks 25X in FIG. 5, that is, marks HM1 and HM2, FIG. 7B is a cross-sectional view of FIG. 7A, and the first mark HM1 is The surface of the wafer 11A is a pattern in which five line-shaped concave patterns 33a having a line width e are formed in a grid pattern with a pitch P2 in the X direction at a predetermined step Hh, and the pitch P2 is 5 to 5. It is about 20 μm. On the other hand, the second mark HM2 is a pattern in which the outer periphery thereof is surrounded by the engraved portion 34, and five elongated linear convex patterns 35a having a line width e are formed in a lattice shape at a pitch P2.

  And the design value of the space | interval to the X direction of the center of both marks HM1 and HM2 is Dh, and the space | interval Dh is about 50-100 micrometers. In this case, the first mark HM1 is a mark in which the width of the concave pattern 33a is e and the width of the convex pattern 33b is f (= P2-e), and the second mark HM2 is the width of the convex pattern 35a is e. The concave pattern 35b is a mark having a width f, and both marks HM1 and HM2 are marks in which only the concavo-convex relationship is reversed. That is, the duty ratio of the concave pattern 33a of the first mark HM1 and the duty ratio of the convex pattern 35a of the second mark HM2 are set to be equal.

The step Hh between the marks MH1 and MH2 is preferably about 40 to 100 nm. This is to obtain an image signal with good contrast. Furthermore, in order to obtain an image signal SH with better contrast, the step Hh is desirably about 40 to 60 nm.
The duty ratio (100 × e / P2 (%)) of the concave pattern 33a of the first mark HM1 and the duty ratio (100 × e / P2 (%)) of the convex pattern 35a of the second mark HM2 are 10%. It is desirable that the degree. This is also because if the duty ratio is too small, the contrast of the mark image is lowered and the reproducibility of the position detection result is deteriorated. If the duty ratio is too large, relative to the image of the concave pattern 33a of the first mark HM1 and the image of the convex pattern 35a of the second mark HM2 due to the displacement of the σ stop 3 (due to the adjustment residual of the illumination system). This is because the amount of change in the amount of misalignment becomes small, and the adjustment sensitivity decreases.

  When the two marks HM1 and HM2 arranged close to each other as described above are observed by the position detection device of FIG. 1, image signals obtained by reading out these images in the X direction by the image sensor 22 (or The signal averaged in the non-measurement direction) is defined as an image signal SH in FIG. In the image signal SH of FIG. 7C, in the portions IH1 and IH2 corresponding to the first mark HM1 and the second mark HM2, the mark portions are dark portions. 1 detects the position Xh1 of the center of the portion IH1 and the position Xh2 of the center of the portion IH2 of the image signal SH, and obtains the interval Mh1. The interval Mh1 is a value obtained by multiplying the interval on the imaging surface by 1 / α using the magnification α from the surface of the wafer 11A to the imaging surface in FIG.

Next, the other pair of marks 26X on the X axis in FIG. 5 is moved to the vicinity of the center of the observation visual field of the position detection apparatus in FIG. 1, and the image signal of the image sensor 22 is processed in the same manner as in FIG. Thus, a distance Mh2 is obtained by converting the distance in the X direction between the center of the image of the third mark HM3 and the center of the image of the fourth mark HM4 into a length on the wafer 11A.
Thereafter, the control calculation system 23 of FIG. 1 calculates a deviation amount δMX from the ideal state of the measured value of the interval between the two pairs of marks 25X and 26X as follows. This deviation amount δMX corresponds to a part of the TIS (Tool Induced Shift) of the illumination system of the position detection device of FIG.

δMX = (Mh1−Mh2) / 2 (23)
In a state where there is no adjustment residual in the illumination system, the deviation amount δMX thus determined should be almost zero, but if there is an adjustment residual in the illumination system, two marks that are observed Since the shift amount of the image positions of HM1 and HM2 varies depending on the level difference of each mark, the measured shift amount δMX increases beyond the allowable range.

  Specifically, in this example, the sign and size of the shift amount δMX change according to the position of the σ stop 3 of the illumination system in FIG. Therefore, the control calculation system 23 in FIG. 1 adjusts the position of the σ stop 3 via the σ stop position adjusting mechanisms 5a and 5b so that the absolute value of the shift amount δMX becomes small. Thereafter, the distances Mh1 and Mh2 between the images of the two sets of marks 25X and 26X are measured again, and these are substituted into the equation (23) to obtain a deviation amount δMX. Until the deviation amount δMX falls within the allowable range. By repeating the interval measurement and the position adjustment of the σ stop 3, the position adjustment of the σ stop 3 is completed.

Similarly to the adjustment in the X direction, the distance between the two images of the Y-axis two pairs of marks 25Y and 26Y on the wafer 11A in FIG. 5 is measured, and the difference between these distances is obtained. The adjustment residual of the illumination system with respect to the direction can also be adjusted.
In the above embodiment, the position adjustment of the coma aberration correcting optical system 16 and the position adjustment of the σ stop 3 in FIG. 1 may be performed independently. However, when the position of the σ stop 3 is adjusted, even if coma remains in the detection optical system (imaging optical system), the adjustment can be performed without being affected by the coma. It is efficient to adjust the coma aberration by moving the coma correction optical system 16 after adjusting the position of the σ stop 3 first.

  In this example, the coma correction optical system 16 in FIG. 1 is adjusted to adjust and remove the remaining coma aberration. However, the present invention is not limited to this, and other optical elements such as the objective lens group 10 and the half prism 9 are used. Residual coma and / or other aberrations (particularly asymmetric aberration) may be adjusted and removed by adjusting the position or rotation angle of the member. Further, when adjusting the illumination state, not only the position of the σ stop 3 but also the position of the light source 1 or the position or rotation angle of the first relay lens 6 or the second relay lens 8 may be adjusted. Good.

Next, another example of the test mark that can be used when adjusting the detection optical system and the illumination system will be described with reference to FIGS.
First, FIG. 8A shows a case where a pair of marks having a rotation angle of 0 ° is formed in the vicinity of the two pairs of marks 25A and 26X on the X axis in FIG. In FIG. 8A, a set of marks 25XA and 26X has the same shape and positional relationship as the marks 25X and 26X in FIG. 5, respectively, and the first mark HM1 and the first mark are located in the vicinity of these marks. A pair of marks 25XB (marks with a rotation angle of 0 °) including the fifth mark HM5 and the sixth mark HM6 having the same shape as the two marks HM2 and the positional relationship are formed. In this example, for example, the distance Mh1 in the left end mark 25XA and the distance Mh2 in the center mark 26X are simultaneously measured in the observation field of view of the position detection device in FIG. In the observation field of view, the distance Mh3 in the rightmost mark 25XB and the distance Mh2 in the center mark 26X are measured at the same time, and Mh3 is substituted for Mh1 in the equation (23). By substituting, a deviation amount δMX (this is assumed to be δMX2) is obtained. Then, for example, by newly setting the shift amount obtained by averaging these shift amounts δMX1 and δMX2 as the shift amount δMX, it is possible to reduce the influence of variations in the detection result due to the position in the observation field of view.

Also in this example, as shown in FIG. 8B, the three pairs of marks 25XA, 26X, and 25XB may be shifted in the non-measurement direction (Y direction).
Next, FIG. 9 shows an example of a mark suitable for use in adjusting the asymmetric aberration of the detection optical system, for example. In FIG. 9, the surface of the adjustment wafer 11B has a period in the X direction. A pair of marks 29X on the X axis composed of a first mark DM1 and a second mark DM2 that are formed in a concave / convex pattern and are arranged in series, and a pair of marks 30X that are formed by rotating the mark 29X by 180 ° It is formed in close proximity. That is, if the distance (design value) between the first mark DM1 and the second mark DM2 is Dd, the pair of marks 30X has the same shape as the marks DM1 and DM2, and the distance (design value) is a third value Dd. It consists of a mark DM3 and a fourth mark DM4.

Further, two pairs of marks 29Y and 30Y on the Y axis are formed by rotating the two pairs of marks 29X and 30X by 90 °. Also in this example, instead of arranging in the X direction as two pairs of X-axis marks 29X and 30X, it is arranged in the Y direction as shown by two pairs of X-axis marks 29X 'and 30X'. Also good.
Also in this example, first, one pair of marks 29X on the X axis in FIG. 9 is moved within the observation field of view of the position detection apparatus in FIG.

  10A is an enlarged plan view showing one pair of marks 29X on the X axis shown in FIG. 9, FIG. 10B is a sectional view of FIG. 10A, and the first mark DM1 is a wafer. 11B is a pattern in which five linear concave patterns 31a having a line width a are formed in a lattice pattern with a pitch P in the X direction at a predetermined step Hd, and the pitch P is about 5 to 20 μm. It is. Similarly, the second mark DM2 is a pattern in which five linear concave patterns 32a having a line width c are formed on the surface of the wafer 11B in a lattice pattern at the same pitch P in the X direction at the level difference Hd. is there. In the X direction that is the measurement direction, the distance between the center of the first mark DM1 and the center of the second mark DM2 is set to Dd as a design value. This distance Dd is about 50 to 100 μm.

  The step (depth) Hd of each of the marks DM1 and DM2 is preferably about 40 to 100 nm. If the level difference Hd is too small, the contrast of those mark images is lowered (the mark portion is not sufficiently dark), and the position detection accuracy is lowered. On the other hand, when the step Hd is higher than about 100 nm, adverse effects such as geometric optical vignetting due to the step portion occur, and high-precision measurement becomes difficult. Further, in order to obtain an image with good contrast for each mark, the step Hd is desirably 40 to 60 nm.

  In this example, the ratio (a: b or a / b) between the width a of the concave pattern 31a of the first mark DM1 and the width b (a + b = P) of the convex pattern 31b, and the concave pattern of the second mark DM2. The ratio (c: d or c / d) of the width c of 32a and the width d (c + d = P) of the convex pattern 32b is different. In other words, the duty ratio (100 × a / P (%)) of the width a of the concave pattern 31a of the first mark DM1 to the pitch P and the duty ratio of the width c of the concave pattern 32a of the second mark DM2 to the pitch P ( 100 × c / P (%)). As an example, in this example, the duty ratio of the concave pattern 31a of the first mark DM1 is set to 50%, and the ratio between the width a of the concave pattern 31a and the width b of the convex pattern 31b is set as follows. Yes.

a: b = 1: 1 or a / b = 1 (24)
On the other hand, the duty ratio of the concave pattern 32a of the second mark DM2 is set to about 10%, and the ratio between the width c of the concave pattern 32a and the width d of the convex pattern 32b is set as follows.
c: d = 1: 9, or c / d = 1/9 (25)
When the two marks DM1 and DM2 arranged close to each other in the X direction are observed with the position detection device of FIG. 1, image signals obtained by reading out the images in the X direction by the image sensor 22 are obtained. Assume that the image signal SD in FIG. The image signal SD may be an image signal obtained by scanning the two mark images of FIG. 10A in the X direction and averaged in the non-measurement direction (Y direction).

  10C, the horizontal axis represents the position X in the measurement direction, and the position X actually uses the magnification α from the surface of the wafer 11A in FIG. 9 to the imaging surface of the imaging element 22 in FIG. The position obtained by multiplying the position from the predetermined reference point on the imaging surface of the imaging element 22 by 1 / α. In the image signal SD of FIG. 10C, in the portion ID1 corresponding to the image of the first mark DM1, the edge portion of the first mark DM1 is a dark portion, and similarly in the portion ID2 corresponding to the image of the second mark DM2. The edge part is a dark part.

  The control calculation system 23 in FIG. 1 obtains an interval Md1 between the center position Xd1 of the part ID1 and the center position Xd2 of the part ID2 from the image signal SD. This distance Md1 should be the designed distance Dd in FIG. 10A when there is no aberration in the detection optical system. However, if coma aberration remains in the detection optical system, the amount of change in image position due to coma aberration varies depending on the duty ratio of the widths of the concave patterns 31a and 32a of both marks DM1 and DM2. Md1 does not match Dd.

  Next, the other pair of marks 30X on the X axis in FIG. 9 is moved into the observation field of view of the position detection apparatus in FIG. 1, and the distance between the images of the marks DM3 and DM4 is converted into the length on the wafer. Obtain Md2. Thereafter, the coma aberration is adjusted by adjusting the position of the coma correction optical system 16 in FIG. 1 so that the two intervals Md1 and Md2 match within a predetermined allowable range. Similarly, adjustment of coma aberration in the Y-axis direction is also performed.

In the above embodiment, the mark rotated by 180 ° is formed in advance on the adjustment wafer, but another pair of marks rotated by a predetermined angle θ (0 ° <θ <360 °). These marks may be formed in advance on the wafer, and the interval between these two pairs of marks may be measured.

Furthermore, depending on the type of aberration to be measured, for example, a pair of first marks and a pair of second marks obtained by rotating the first marks by 180 °, and a pair of third marks obtained by rotating the pair of first marks by 45 °. A mark and a pair of fourth marks obtained by rotating the marks 180 ° are formed on the wafer, and the imaging optical system and illumination are also used by using the intervals detected by the pair of third marks and fourth marks, respectively. The system may be adjusted. At this time, a pair of fifth marks obtained by rotating the pair of first marks by 135 ° and a pair of sixth marks obtained by rotating the first marks by 180 ° may be further formed on the wafer. By using each of the four pairs of the third to sixth marks, it is possible to detect the characteristics of each component in the sagittal direction (S direction) and the meridional direction (M direction) for the optical characteristics to be measured. Become.

  In the second embodiment, for example, a pair of first marks and a pair of second marks obtained by rotating the first marks by a predetermined angle (for example, 180 °) are formed on the wafer. And the pair of second marks, that is, the actual pair of second marks (in the arrangement direction) when the pair of first marks is precisely rotated by the predetermined angle (the arrangement direction). The remaining rotation error in the direction of the arrangement may be measured. When the measured remaining rotation error (difference between the actual rotation angle and the predetermined angle) is larger than a predetermined value, for example, the interval detected by the pair of second marks is set according to the remaining rotation error. You may make it correct | amend. Thereby, the space | interval of a pair of 2nd mark rotated only the predetermined angle correctly on the basis of a pair of 1st mark can be obtained.

In the above embodiment, the adjustment marks are formed on the adjustment wafer. However, these marks may be formed on a reference plate such as an XY stage.

[Third Embodiment]
Next, a third embodiment of the present invention will be described with reference to FIGS. 1 and 11 to 13. Since the position detection apparatus of FIG. 1 is also the object of adjustment in this example, description of the apparatus configuration is omitted, and an adjustment method of the optical system will be described. In this example, a wafer 11C is installed instead of the adjustment wafer 11 shown in FIG. A plurality of evaluation marks (step marks) are formed on the surface of the wafer 11C. First, a description will be given of evaluation marks made up of a plurality of sets of concave and convex lattice marks (step marks) formed on the adjustment wafer 11C of this example.

  FIG. 11 is a plan view showing an adjustment wafer 11C installed instead of the wafer 11 in FIG. 1. In FIG. 11, a silicon wafer is used as an example of the wafer 11C. On the surface of the wafer 11C, an X-axis first evaluation mark DX made up of a first mark DM21 to a third mark DM23 each having a periodic uneven pattern in the X direction, and the evaluation mark DX are placed at 90 °. A first evaluation mark DY on the Y axis composed of the first mark DM31 to the third mark DM33 having a rotated shape is formed. The X-axis evaluation mark DX is a mark that is substantially line-symmetric in the measurement direction, in which marks DM21 and DM23 having different width ratios between the concave and convex portions are arranged at both ends of the central mark DM22. The evaluation marks DX and DY are used for adjusting the characteristics of the detection optical system.

  Further, on the surface of the wafer 11C, an X-axis second evaluation mark HX composed of a first mark HM21 to a third mark HM23 each having a periodic uneven pattern in the X direction, and this evaluation mark HX are provided. A second evaluation mark HY on the Y axis made up of the first mark HM31 to the third mark HM33 that is rotated by 90 ° is formed. The X-axis evaluation mark HX is a substantially line-symmetrical mark in the measurement direction in which marks HM21 and HM23 having a shape in which the relationship between the concave portion and the convex portion is reversed are arranged on both ends of the central mark HM22. The evaluation marks HX and HY are used for adjusting the characteristics of the illumination system.

For example, when a silicon wafer is used as the wafer 11C, these concave and convex lattice marks are formed by applying a photoresist onto the surface, exposing a projected image of the corresponding reticle pattern, developing the photoresist, etching, and resist. It can be formed with extremely high accuracy by a process such as peeling.
12A is an enlarged plan view showing the first evaluation mark DX on the X axis shown in FIG. 11, FIG. 12B is a sectional view of FIG. 12A, and the mark DM22 in the center is This is a pattern in which three linear concave patterns 42c having a line width c are formed in a lattice pattern with a pitch P in the X direction at a predetermined step Hd on the surface of the wafer 11C, and the pitch P is about 5 to 20 μm. It is. Further, marks DM21 and DM23 are formed so as to sandwich the mark DM22. Similarly, the mark DM21 has three linear concave patterns 41a having a line width a formed on the surface of the wafer 11C with a step Hd in the X direction. Are patterns formed in a lattice pattern at the same pitch P. The shape of the mark DM23 is the same as that of the mark DM21.

  In the measurement direction (X direction), the distance between the center of the mark DM21 and the center of the mark DM22 and the distance between the center of the mark DM22 and the center of the mark DM23 are set to Dd as a design value. This interval Dd is about 40 to 60 μm. As a result, the center of the mark DM22 coincides with the centers of the marks DM21 and DM23 on both sides in the measurement direction by design.

  Further, the step (depth) Hd of each of the marks DM21 to DM23 is preferably about 50 to 100 nm. If the level difference Hd is too small, the contrast of those mark images is lowered (the mark portion is not sufficiently dark), and the position detection accuracy is lowered. On the other hand, when the step Hd is higher than about 100 nm, adverse effects such as geometric optical vignetting due to the step portion occur, and high-precision measurement becomes difficult.

  In this example, the ratio of the width a of the concave pattern 41a of the mark DM21 at one end (the same applies to the mark DM23 at the other end) to the width b (a + b = P) of the convex pattern 41b (a: b, Or a / b) is different from the ratio (c: d or c / d) of the width c of the concave pattern 42c of the center mark DM22 and the width d (c + d = P) of the convex pattern 42d. As an example, in this example, the width a of the concave pattern of the marks DM21 and DM23 is set to 50% of the pitch P, and the ratio of the width a of the concave pattern to the width b of the convex pattern is set as follows. Yes.

a: b = 1: 1 or a / b = 1 (31)
On the other hand, the width c of the concave pattern of the center mark DM22 is set to about 10% of the pitch P, and the ratio of the width c of the concave pattern to the width d of the convex pattern is set as follows.
c: d = 1: 9, or c / d = 1/9 (32)
When the evaluation marks composed of the three marks DM21, DM22, and DM23 arranged close to each other in the X direction are observed with the position detection apparatus shown in FIG. The image signal obtained by reading is assumed to be an image signal SD in FIG. The image signal SD may be an image signal obtained by scanning the image of the three marks in FIG. 12A in the X direction and averaging it in the non-measurement direction (Y direction).

  In FIG. 12C, the horizontal axis represents the position X in the measurement direction, and the position X is actually imaged using the magnification α from the surface of the wafer 11C in FIG. This represents a position obtained by multiplying the position from a predetermined reference point on the imaging surface of the element 22 by 1 / α. In the image signal SD of FIG. 12C, each edge portion is a dark portion in the portions ID21 and ID23 corresponding to the images of the marks DM21 and DM23 at both ends, and a narrow width is in the portion ID22 corresponding to the image of the center mark DM22. The part corresponding to the concave pattern c is a dark part.

The control calculation system 23 in FIG. 1 obtains the center positions Xd21, Xd22, and Xd23 of the partial IDs 21, 22, and ID23 from the image signal SD as the positions of the marks, and then calculates the average position Xd24 of the marks at both ends. Calculate as follows.
Xd24 = (Xd21 + Xd23) / 2 (33)
Next, the control calculation system 23 calculates the deviation Md (= Xd22−Xd24) of the average position Xd24 of the marks DM21 and DM23 at both ends with respect to the position Xd22 of the center mark DM22 of the evaluation mark as a relative positional relationship.

  The deviation Md measured in this way should be a predetermined reference value D0 (0 in design) obtained in a state where there is no aberration in the detection optical system. However, if coma aberration remains in the detection optical system, an image due to coma aberration is caused by a difference in duty ratio between the mark DM22 and the marks DM21 and DM23 (ratio of the concave pattern widths c and a to the pitch P). Since the amount of change in position is different, the measured deviation Md does not match the reference value D0. Further, not only the magnitude of the remaining coma aberration but also the sign can be determined from the magnitude relationship between the measured deviation Md and the reference value D0. Therefore, an error ΔMd (= D0−Md) of the deviation Md with respect to the reference value D0. 1) adjusts the position of the frame correction optical system 16 via the frame correction optical system position adjustment mechanisms 17a and 17b so that the error ΔMd becomes small. .

  After adjusting the position of the frame correction optical system 16, the deviation Md between the position of the center mark DM22 and the average position of the marks DM21 and DM23 at both ends is measured again, and the error ΔMd of the deviation Md with respect to the reference value D0 is measured. Is calculated. If the error ΔMd is out of the allowable range, the above adjustment is performed again. By repeating the above steps until the error ΔMd is driven within the allowable range, the adjustment of the coma aberration of the detection optical system is completed. Since the adjustment method of this example is a method of measuring mark intervals having different duty ratios as in the first embodiment, it will be referred to as a “different ratio mark method”.

  Note that there are various methods such as a slice method and a correlation method in the position detection algorithm of each mark image in the above-described position detection process, and any of them may be used in this example. For example, in the slice method, position detection is performed based on the slice position of the image signal SD at a predetermined slice level, and in the correlation method, the image signal SD is compared with a predetermined reference waveform, and the degree of correlation with the reference waveform is determined. The highest position is the mark image position.

  The reference value D0 is supposed to be 0 due to the symmetry of the evaluation mark, but the reticle pattern manufacturing error used when the evaluation mark is transferred, or on the wafer 11C. The actual reference value D0 may not be zero due to a slight asymmetry of the evaluation mark based on a manufacturing error such as an etching error when forming a step and distortion of the detection optical system. Therefore, it is desirable to measure the actual value of the reference value D0 before the above adjustment. For this purpose, for example, as disclosed in the prior art and “Document 1” cited in the first embodiment, the first measurement value and the second measurement value when the wafer 11C is rotated 180 °. The average value may be adopted.

  Specifically, in this example, after measuring the deviation (referred to as Md1) in the X direction of the average position of the marks DM21 and DM23 at both ends with respect to the position of the center mark DM22 as described above, the wafer 11C is rotated by 180 °. Then, the deviation (referred to as Md2) in the X direction of the average position of the marks DM23 and DM21 at both ends with respect to the position of the center mark DM22 is measured again. Then, as represented by the following equation, an average value of these two measured values may be adopted as the reference value D0.

D0 = (Md1 + Md2) / 2 (34)
Thereby, even if there is residual aberration in the detection optical system of FIG. 1, it is possible to detect with high accuracy the error ΔMd of the center deviation Md of the marks at both ends with respect to the center mark due to coma aberration.
However, in such a method of measuring the reference value (true value) by rotating the wafer 11C by 180 °, the evaluation mark itself on the wafer 11C is also rotated by 180 °, and the evaluation mark is symmetrical. Without (line symmetry with respect to the central axis along the non-measurement direction (Y direction) orthogonal to the measurement direction), the measured optical value for the detection of the positional relationship between the two marks before the rotation and after the 180 ° rotation is obtained. There is a possibility that an error component due to system distortion or the like is superimposed. In this regard, in this example, since the evaluation mark DX has the symmetry as shown in FIG. 12, the symmetry of the evaluation mark DX before the rotation and at the time of 180 ° rotation is maintained, and the detection mark DX is detected. It is possible to measure the reference value D0 with high accuracy without including any error component due to distortion of the optical system for use.

  By the way, in general, a position detection device needs to measure a mark position in a two-dimensional direction (X direction, Y direction) or a relative positional relationship. Therefore, similarly to the adjustment in the X direction described above, the positions of the images of the marks DM31 to DM33 of the first evaluation mark DY on the Y axis on the wafer 11C in FIG. 11 are detected, and the image of the image of the center mark DM22 is detected. By measuring the deviation of the center of the image of the marks DM31 and DM33 at both ends with respect to the center as a relative positional relationship, the coma aberration in the Y direction can also be adjusted.

Next, a method for adjusting the illumination system of FIG. 1 in this example will be described. For this purpose, first, the X-axis second evaluation mark HX on the adjustment wafer 11C of FIG. 11 is used.
13A is an enlarged plan view showing the X-axis second evaluation mark HX shown in FIG. 11, FIG. 13B is a cross-sectional view of FIG. 13A, and the center mark HM22 is The outer periphery is surrounded by the engraved portion 34, and three elongated linear convex patterns 44e having a line width e are formed in a lattice shape in the X direction at a pitch P2. The pitch P2 is about 5 to 20 μm.

  Then, the marks HM21 and HM23 having the same shape are formed so as to sandwich the mark HM22 in the X direction. The mark HM21 has two elongated concave patterns 43e having a line width e formed on the surface of the wafer 11C. It is a pattern formed in a grid pattern with a pitch P2 in the X direction at the level difference Hh. That is, the mark HM21 at the end (the same applies to the other mark HM23) is a mark in which the width of the concave pattern 43e is e and the width of the convex pattern 43f is f (= P2-e), and the central mark HM22 is convex. The width of the pattern 44e is e and the width of the concave pattern 44f is f. The marks HM21 and HM23 at both ends and the mark HM22 at the center are marks in which only the concavo-convex relationship is reversed. In other words, the duty ratio of the concave pattern of the marks HM21 and HM23 at both ends and the duty ratio of the convex pattern of the center mark HM22 are set to be equal to each other.

Further, the width of the concave pattern of the marks HM21 and HM23 at both ends, that is, the width e of the convex pattern of the central mark HM22 is set to about 5% to 10% of the mark pitch P2, for example.
In the measurement direction (X direction), the distance between the center of the mark HM21 and the center of the mark HM22 and the distance between the center of the mark HM22 and the center of the mark HM23 are set to Dh as a design value. This distance Dh is about 40 to 60 μm. Also in this case, by design, the center of the mark HM22 coincides with the centers of the marks HM21 and HM23 on both sides in the measurement direction.

  When the three marks HM21, HM22, and HM23 arranged close to each other in this way are observed with the position detection device of FIG. 1, image signals (or obtained by reading out these images in the X direction by the image sensor 22 (or A signal obtained by averaging this in the non-measurement direction) is defined as an image signal SH in FIG. If the width e in FIG. 13A is narrow, in the image signal SH in FIG. 13C, in the portions IH21 and IH23 corresponding to the marks HM21 and HM23 on both sides, the concave pattern portion becomes a dark portion, and the center In the portion IH22 corresponding to the mark HM22, the portion corresponding to the convex pattern is a dark portion.

  1 detects the positions Xh21, Xh22, Xh23 of the centers of the portions IH21, IH22, IH23 of the image signal SH by the slice method or the correlation method, and further detects the average positions Xh24 of the marks at both ends. (= (Xh21 + Xh23) / 2) is calculated. Next, the deviation Mh of the center position Xh24 of the marks HM21 and HM23 at both ends with respect to the position Xh22 of the center mark HM22 is obtained as a relative positional relationship.

In a state where there is no adjustment residual in the illumination system, the deviation Mh thus obtained should be equal to a predetermined reference value H0 (0 in design). Since the shift amount of the position of the image to be changed differs depending on the step (concave or convex) of each mark, the measured deviation Mh is different from the reference value H0.
Specifically, in this example, the sign and the magnitude of the error ΔMh (= H0−Mh) with respect to the reference value H0 of the deviation Mh change according to the position of the σ stop 3 of the illumination system in FIG. Therefore, the control calculation system 23 in FIG. 1 adjusts the position of the σ diaphragm 3 via the σ diaphragm position adjusting mechanisms 5a and 5b so that the error ΔMh becomes small. Thereafter, the error ΔMh of the deviation Mh of the marks at both ends with respect to the center mark is measured again, and the position adjustment of the σ stop 3 is completed by adjusting the position of the σ stop 3 until the error ΔMh falls within the allowable range. To do.

  This adjustment method is referred to as “different step mark method” in the present specification, as it measures the positional relationship between marks of different steps. The different step mark method is a method disclosed in the above-mentioned document 1, but if the inventor of the present application uses the different step mark method, the displacement of the σ stop 3 or the arrangement surface of the σ stop 3 is changed. It has been found that the illuminance non-uniformity of the illumination light can be adjusted with high accuracy.

  Incidentally, the step Hh of the marks HM21, HM22, HM23 is preferably about 30 to 60 nm. This step Hh is within the range or less, and the smaller it is (the lower the step is), the greater the change in the detected positional relationship of each mark due to the displacement of the σ stop 3 or the like. . That is, the detection sensitivity is increased. However, if the mark step Hh is too small, the contrast of the mark image is lowered and the SN ratio of the image signal is lowered, so that the measurement accuracy of the positional relationship is lowered. Therefore, it is desirable that the mark step Hh be in the above range, but if the SN ratio of the image pickup element of the detection optical system to be used is good, the detection sensitivity is improved by using a lower step mark. Needless to say.

  The width e of the concave pattern of the marks HM21 and HM23 on both sides and the width e of the convex pattern of the central mark HM22 are preferably about 5 to 10% of the pitch P2. This is also because if the duty ratio of the width e (ratio to the pitch P2) is too small, the contrast of the mark image is lowered and the reproducibility of the position detection result is deteriorated. On the other hand, if the duty ratio is too large, the concave pattern image of the marks HM21 and HM23 on both sides and the convex pattern image of the center mark HM22 due to the displacement of the σ stop 3 (due to the adjustment residual of the illumination system) are relative. This is because the amount of change in the amount of misalignment becomes small, and the adjustment sensitivity decreases.

Similarly to the adjustment in the X direction, by measuring the positional relationship of the image of the second evaluation mark HY on the Y axis on the wafer 11C in FIG. 11, the adjustment residual of the illumination system in the Y direction is measured. Adjustments can also be made.
Further, also in the case of FIG. 13, it is desirable to measure the actual value of the reference value H0 of the deviation Mh between the marks HM21 and HM23 on both sides of the center mark HM22. For this purpose, as in the case of FIG. 12, after measuring the deviation (referred to as Mh1) of the marks HM21 and HM23 with respect to the mark HM22 in FIG. 13A, the wafer 11C is rotated by 180 ° and similarly to the mark HM22. The deviation (marked as Mh2) of the marks HM23 and HM21 is measured, and the average value of these two measured values may be adopted as the reference value H0.

Note that, as the evaluation mark, a mark having a different configuration from each of the evaluation marks shown in FIGS. 12 and 13 may be used.
For example, in the example of FIG. 12, the ratio of the width of the concave portion to the width of the convex portion is 1: 1 for the center mark DM22, and the duty ratio of the concave portion width is 5 to 10% for the marks DM21 and DM23 at both ends. It is good also as a concave mark of a grade. In addition, a mark (convex mark) having a convex portion width duty ratio of about 5 to 10% may be used instead of the mark having a concave portion width duty ratio of about 5 to 10%.

Similarly, in the example of FIG. 13, the central mark HM22 may be a mark having a narrow concave portion (concave mark), and the marks HM21 and HM23 on both sides may be a mark having a narrow convex portion (convex mark).
Further, the number of concave patterns (or convex patterns) constituting each of the marks DM21, DM22, DM23, HM21, HM22, and HM23 is not limited to the example shown in FIGS. 12 and 13 and may be any number. However, in order to maintain the symmetry of the entire evaluation mark, it is desirable that the number of concave patterns (or convex patterns) of the marks at both ends is equal. That is, it is desirable that the number of concave patterns constituting the marks DM21 and DM23 in FIG. 12 is equal to each other, and the number of concave patterns constituting the marks HM21 and HM23 in FIG.

  In the above embodiment, the position adjustment of the coma aberration correcting optical system 16 and the position adjustment of the σ stop 3 in FIG. 1 may be performed independently. However, when the position of the σ stop 3 is adjusted, even if coma remains in the detection optical system (imaging optical system), the adjustment can be performed without being affected by the coma. It is efficient to adjust the coma aberration by moving the coma correction optical system 16 after adjusting the position of the σ stop 3 first.

  In this example, the coma correction optical system 16 in FIG. 1 is adjusted to adjust and remove the remaining coma aberration. However, the present invention is not limited to this, and other optical elements such as the objective lens group 10 and the half prism 9 are used. The remaining coma aberration may be adjusted and removed by adjusting the position or rotation angle of the member. Further, when adjusting the illumination state, not only the position of the σ stop 3 but also the position of the light source 1 or the position or rotation angle of the first relay lens 6 or the second relay lens 8 may be adjusted. Good.

  In the third embodiment described above, instead of rotating the wafer 11C by 180 °, evaluation marks (DX, DY, HX, HY) including three concave and convex marks are provided in the same manner as in the second embodiment. Another evaluation mark rotated by a predetermined angle (for example, 180 °) may be further formed on the wafer 11C. Thereby, it is possible to obtain the average values (D0, H0) of deviations detected at 0 ° and 180 °, respectively, without rotating the wafer 11C.

  By the way, the number of the concavo-convex marks used in the first to third embodiments described above may be arbitrary. Further, using one set of mark groups composed of a plurality of concave and convex marks having a rotation angle of 0 °, and another set of marks rotated by 90 °, an imaging optical system in the X direction and Y direction, respectively. The optical characteristics of the lighting system and the illumination system are obtained, and two sets of mark groups are formed on the wafer by rotating one set of mark groups whose rotation angle is 0 ° by 45 ° and 135 °, respectively. Optical characteristics and the like may be obtained in the (S direction) and meridional direction (M direction), respectively.

Next, a case where the position detection apparatus of FIG. 1 is applied to an alignment sensor of a projection exposure apparatus will be described with reference to FIGS.
FIG. 14 shows a projection exposure apparatus used in this example. In FIG. 14, during exposure, an illumination light source such as a mercury lamp or an excimer laser light source, an optical integrator, a variable field stop, a condenser lens system, etc. The optical system 51 irradiates the reticle R with exposure light IL. Then, an image of the pattern formed on the reticle R is one shot on the wafer W coated with a photoresist at a projection magnification β (β is 1/5, 1/4, etc.) via the projection optical system PL. Projected into the area. At this time, the exposure amount control system 52 optimizes the exposure amount based on the control information of the main control system 53.

  Hereinafter, taking the Z axis parallel to the optical axis AX of the projection optical system PL, taking the X axis parallel to the paper surface of FIG. 14 in the plane perpendicular to the Z axis, and taking the Y axis perpendicular to the paper surface of FIG. explain. At this time, the reticle R is sucked and held on the reticle stage 54, and the reticle stage 54 is placed on the reticle base 55 based on control information of the drive system 57 based on the measurement value of the coordinates of the reticle stage 54 by the laser interferometer 56. To position the reticle R in the X direction, the Y direction, and the rotation direction.

  On the other hand, the wafer W is held on a wafer holder (not shown) by vacuum suction, the wafer holder is fixed on the sample table 58, and the sample table 58 is supported so as to float on the surface plate 60 via an air bearing. It is fixed on the XY stage 59. The sample stage 58 controls the position (focus position) and the tilt angle of the wafer W in the Z direction to align the surface of the wafer W with the image plane of the projection optical system PL by an autofocus method, and the XY stage 59 performs laser interference. Based on the control information of the drive system 62 based on the position of the sample stage 58 measured by the meter 61, the sample stage 58 is moved stepwise in the X direction and the Y direction. By repeating the step movement by the XY stage 59 and the exposure of the pattern image of the reticle R by the step-and-repeat method, each shot area on the wafer W is exposed.

  When performing overlay exposure with the projection exposure apparatus of FIG. 14, it is necessary to align the reticle R and the wafer W in advance before the exposure. Therefore, a reference mark member 65 in which various reference marks are formed on the sample stage 65 is fixed, and the reticle R is based on the measurement result of a reticle alignment microscope (not shown) on the reticle R. Are aligned. Further, an image processing type alignment sensor 63 is disposed on the side surface of the projection optical system PL by an off-axis method having the same configuration as the optical system of the position detection device of FIG. 1) is supplied to the alignment signal processing system 64. In addition to the function of the control arithmetic system 23 in FIG. 1, the alignment signal processing system 64 obtains the amount of positional deviation of the image of the alignment mark (wafer mark) on the detection target wafer W with respect to the index marks 14a and 14b in FIG. It has a function.

  The image signal processing method in the alignment signal processing system 64 is disclosed in, for example, Japanese Patent Application Laid-Open No. 4-65603 and US Pat. No. 5,493,403 corresponding thereto. Detailed description is omitted. As long as the national laws of the designated country designated in this international application or the selected country selected allow, the disclosure of this gazette and US patent is incorporated into the text.

  Further, as shown in FIG. 15A, a two-dimensional wafer mark 38 for alignment is formed in each shot area 36 on the wafer W, and the wafer mark 38 has an uneven pattern at a constant pitch in the Y direction. The Y-axis wafer mark 37 </ b> Y is formed on the X-axis, and the X-axis wafer mark 37 </ b> X is formed with an uneven pattern formed in the X direction at a constant pitch so as to sandwich the wafer mark 37 </ b> Y. Two or more wafer marks 38 may be formed for one shot area. Further, as shown in FIG. 15B, the wafer mark to be detected may be one-dimensional wafer marks 40X and 40Y that are independently attached to each shot area on the wafer. The former X-axis wafer mark 40X is a mark formed with an uneven pattern at a constant pitch in the X direction, and the latter Y-axis wafer mark 40Y is a mark formed with an uneven pattern at a constant pitch in the Y direction. is there.

  When the alignment sensor 63 is used in FIG. 14, the adjustment wafer 11 shown in FIG. 2 is first placed on the sample stage 58 via a wafer loader system (not shown). Thereafter, the distance between the images of the two pairs of marks (HM1, HM2) and marks (HM11, HM12) for adjusting the characteristics of the illumination system of FIG. 2 is measured via the alignment sensor 63, and the figure is based on the measurement result. The position of 1 σ stop 3 is adjusted. Thereafter, the distance between the images of the two pairs of marks (DM1, DM2) and marks (DM11, DM12) for adjusting the characteristics of the optical system for detection shown in FIG. Based on this, the position of the frame correction optical system 16 in FIG. 1 is adjusted.

  The same adjustment can be performed using the adjustment wafer 11C shown in FIG. In this case, the adjustment wafer 11C shown in FIG. 11 is first placed on the sample stage 58 via a wafer loader system (not shown). Thereafter, the positional relationship between the images of the two evaluation marks HX and XY for adjusting the characteristics of the illumination system in FIG. 11 is measured via the alignment sensor 63, and the position of the σ stop 3 in FIG. Adjust. Thereafter, the positional relationship between the images of the two evaluation marks DX and DY for adjusting the characteristics of the detection optical system shown in FIG. 11 is measured via the alignment sensor 63, and the frame correction optics shown in FIG. The position of the system 16 may be adjusted. Furthermore, the illumination system can be adjusted by measuring the interval between the images of the marks 25X, 26X, 25Y, and 26Y on the adjustment wafer 11A shown in FIG. 5, and the marks 28A and 28B on the wafer 11A shown in FIG. The optical system for detection can be adjusted by measuring the positional relationship of the images or the interval between the images of the marks 29X, 30X, 29Y, and 30Y on the wafer 11B in FIG.

  Thereafter, in FIG. 14, in parallel with the alignment of the reticle R, the position of a predetermined reference mark on the reference mark member 65 is detected via the alignment sensor 63, whereby the center position (exposure center) of the pattern image of the reticle R is detected. ) And the detection center of the alignment sensor 63 (baseline amount) is obtained. Thereafter, by driving the XY stage 59 based on coordinates obtained by correcting the position of the wafer mark detected via the alignment sensor 63 with the baseline amount, high overlay accuracy can be obtained.

  The configuration of the reference mark member 65, the reticle alignment, the measurement of the baseline amount, and the like are disclosed in, for example, Japanese Patent Laid-Open No. 4-324923 and US Pat. No. 5,243,195 corresponding thereto. Therefore, to the extent permitted by the national laws of the designated country designated in this international application or the selected country of choice, the disclosure of this gazette and US patent is incorporated into the text.

  Further, the σ stop 3 and the stop 18 shown in FIG. 1 are configured to be exchangeable with different stops. For example, a plurality of stops are provided on the turret plate in the illumination system and the imaging optical system, Depends on the formation conditions (level difference, pitch, shape, etc.) of the alignment mark (for example, wafer mark 38), the type of layer (reflectance, etc.) on the wafer W on which the alignment mark is formed, the type of resist, film thickness, etc. One stop selected in this manner may be arranged in the optical path. For example, in order to detect the wafer mark 38 by the dark field method, the diaphragm 18 is replaced with a diaphragm that shields zero-order light (regular reflection light) generated from the wafer, or has a function similar to that of a phase contrast microscope. Therefore, the diaphragm 18 may be replaced with a phase difference plate. Further, simultaneously or independently with the replacement of the diaphragm 18, the σ diaphragm 3 may be replaced with a diaphragm having a ring-shaped opening so that the wafer mark 38 is illuminated with a ring. In the alignment sensor 63 having such a configuration, when at least one of the illumination system and the imaging optical system is replaced, for example, the wafer 11 shown in FIG. It is desirable to adjust the optical system. An alignment sensor that can exchange the diaphragm in the illumination system and the imaging optical system is disclosed in, for example, Japanese Patent Laid-Open No. 8-306609 and Japanese Patent Laid-Open No. 8-327318, and US Pat. No. 5 corresponding thereto. , 706, 091, and to the extent permitted by the national laws of the designated country designated in this international application or the selected country of choice, the disclosure of these gazettes and US patents is incorporated to the fullest extent. Part.

  Further, when any one of the wafers 11 and 11A to 11C is loaded on the wafer stage 59 in the projection exposure apparatus shown in FIG. It is desirable to minimize the rotation error of the wafer with respect to the coordinate system (that is, the orthogonal coordinate system defined by the interferometer 61). Therefore, a notch (notch or the like) of the wafer and other peripheral portions are detected, and a positional deviation and a rotation error of the wafer W in the X direction and the Y direction are obtained, and these measured values are almost 0 (zero). In this manner, the wafer and the wafer stage are moved relative to each other and then held by suction with the wafer holder. A wafer pre-alignment mechanism is disclosed in, for example, Japanese Patent Application Laid-Open No. 7-288276 and US Application No. 391,648 (application date: February 21, 1995) corresponding thereto. As long as the national laws of the designated country specified in the application or the selected selected country permit, the above publications and the disclosure of the US application will be incorporated into the text.

  In the projection exposure apparatus of FIG. 14, the position detection apparatus of FIG. 1 is used as an off-axis type alignment sensor. The alignment sensor used in the projection exposure apparatus is on the wafer via the projection optical system PL. A TTL (through-the-lens) method for detecting a mark or a TTR (through-the-reticle) method for detecting a mark on a reticle and a mark on a wafer may be used. Although not shown in FIG. 14, a large number of optical elements constituting the off-axis alignment sensor 63 are separately held in a plurality of lens barrels, and a gantry on which the projection optical system PL is placed. Each lens barrel is fixed to an integrally provided hardware.

  Further, the projection exposure apparatus in FIG. 14 is not limited to the step-and-repeat method, and the projection exposure apparatus in FIG. 14 is applied to a scanning exposure method such as a step-and-scan method or a mirror projection method, or a photosensitive substrate. A step-and-stitch method in which a plurality of patterns are partially overlapped and transferred may be used. In addition, a soft X-ray region (wavelength of about 5 to 15 nm) generated from a laser plasma light source or SOR (Synchrotron Orbital Radiation) ring as exposure illumination light, for example, EUV (Extreme UltraViolet) having a wavelength of 13.4 nm or 11.5 nm You may comprise as a reduction projection type exposure apparatus (EUV exposure apparatus) using light, a proximity type X-ray exposure apparatus using hard X-rays, or an exposure apparatus using charged particle beams such as an electron beam or an ion beam. In the EUV exposure apparatus, the reduction projection optical system is a reflection system composed of only a plurality of (about 3 to 6) reflection optical elements, and a reflective reticle is used as the reticle.

Further, a reticle or mask used in an exposure apparatus for manufacturing a device for manufacturing a semiconductor element or the like may be manufactured by an exposure apparatus using, for example, far ultraviolet light or vacuum ultraviolet light, and the present invention manufactures a reticle or mask. Therefore, the present invention can also be applied to an exposure apparatus used in a lithography process.
Further, g-line or i-line of a mercury lamp, laser light such as KrF excimer laser light, ArF excimer laser light or F ₂ laser light, or harmonics of YAG laser may be used as exposure illumination light. Alternatively, a DFB (Distributed Feedback) semiconductor laser, for example, is amplified by a fiber amplifier doped with erbium (Er) (or both erbium and ytterbium (Yb)) as exposure illumination light, and a nonlinear optical crystal is used. Alternatively, harmonics converted to ultraviolet light may be used.

  Further, the projection optical system PL in FIG. 14 may be any of a refraction system, a reflection system, and a catadioptric system. As the catadioptric system, for example, as disclosed in US Pat. No. 5,788,229, a plurality of refractive optical systems and two reflective optical elements (at least one is a concave mirror) are arranged in a straight line without being bent. An optical system arranged on the extending optical axis can be used. And, as far as the national laws of the designated country designated in this international application or the selected selected country permit, the disclosure of this US patent is incorporated into the text.

  Next, FIG. 16 shows an example in which the position detection device of FIG. 1 is applied to an overlay error measurement device. In this overlay error measuring device, for example, a relative positional relationship between each mark in a pair of lattice marks or a so-called pair of box-in-box marks may be measured, and the absolute position of each mark is measured. Since it is unnecessary, a highly accurate position measuring device such as a laser interferometer is unnecessary. Then, the wafer W on which the circuit pattern (or resist pattern) is formed on the two layers to be compared is sucked and held by the sample stage 71, and the sample stage 71 is orthogonal to the XY stage 72 2. The two-dimensional position of the sample stage 71 is measured by a linear encoder (not shown), and the measured value is supplied to the control calculation system 73. The control arithmetic system 73 positions the sample stage 71 via the XY stage 72 based on the measured value. The sample table 71 also has a function of adjusting the height (focus position) of the wafer W as a test object in a minute range.

  A position detection device 74 having the same configuration as the optical system of the position detection device in FIG. 1 is disposed above the wafer W, and an image signal of the image sensor 22 of the position detection device 74 is supplied to the control calculation system 73. In addition to the function of the control calculation system 23 of FIG. 1, the control calculation system 73 also has a function of obtaining the positional deviation amount of the two wafer mark images to be detected. Also in this overlay error measuring apparatus, the adjustment wafer 11 of FIG. 2 or the adjustment wafer 11C of FIG. 11 is placed on the sample stage 71 instead of the wafer W, as in the above embodiment. By measuring the distance between the images of each pair of marks, the illumination system and the detection optical system can be adjusted with high accuracy. Thereafter, the wafer W is placed on the sample stage 71, and the control calculation system 73 drives the XY stage 72 to send a pair of marks whose relative positions are to be measured under the position detection device 74, thereby overlapping the positions. The alignment error can be measured with high accuracy.

  In the overlay error measuring apparatus, since two wafer marks to be detected are formed on different layers, the position of the position detecting device 74 in the optical axis direction may be different between the two marks. Therefore, when each of the two wafer marks is detected, depending on the difference between the positions of the two wafer marks in the optical axis direction so that each mark image is accurately focused on the light receiving surface of the image sensor 22 in FIG. For example, it is desirable to move the sample stage 71 in the optical axis direction of the objective lens group 10 or to move the image sensor 22 in the optical axis direction of the imaging optical system.

  Further, the projection exposure apparatus in FIG. 14 or the overlay error measurement apparatus in FIG. 16 detects the position of the focus position of the wafer W (the optical axis direction of the alignment sensor 63 or the optical axis direction of the position detection device 74). In general, an autofocus mechanism for focusing the surface of the wafer W on the alignment sensor 63 or the best focus position of the position detection device 74 is provided. In this case, when detecting the position of the wafer mark on the wafer W or the interval between the pair of wafer marks, the auto-focus mechanism is operated to perform measurement while focusing on the test mark. Similarly, during the adjustment shown in the above embodiment, that is, during the interval measurement of each mark on the wafer 11 shown in FIG. 2, for example, the autofocus mechanism is operated to perform various measurements in a focused state. It is desirable.

  Note that the autofocus mechanism is a TTL system that irradiates a detection beam onto the wafer through the objective lens group 10, or tilts the detection beam on the wafer by tilting the optical axis and the wafer surface without passing through the objective lens group 10. Any of the oblique incident light systems for irradiating the light beam may be used. An alignment sensor having a TTL focus position detection system is disclosed in, for example, Japanese Patent Laid-Open No. 7-321030 and US Pat. No. 5,721,605 corresponding thereto, and is designated as specified in this international application. To the extent permitted by national laws of the country or selected country of choice, the disclosures of the above publications and US patents are incorporated to form part of the text. In the above publication and US patent, the mark on the wafer is irradiated with the above-described detection beam. For example, a street line (scribe line on the wafer) other than the area where the mark is formed on the wafer. ) May be irradiated with a detection beam.

  Further, the alignment sensor 63 and the position detection device 74 irradiate the characteristic adjustment marks formed on the wafers 11, 11 </ b> A to 11 </ b> C or the alignment marks on the wafer W with broadband illumination light and are reflected by the wafers. The indicator marks 14a and 14b are illuminated with the light. However, an illumination system for the index mark may be provided separately from the illumination system that illuminates the mark on the wafer. An alignment sensor having an illumination system for index marks is disclosed in, for example, Japanese Patent Application Laid-Open No. 4-273246 and Japanese Patent Application Laid-Open No. 5-41343, and US Application No. 841,833 (application date) corresponding thereto. : February 26, 1992), and the description in the main text is incorporated with the disclosure of the above publication and US application to the extent permitted by the national laws of the designated country designated in this international application or the selected country of choice. Part.

  Furthermore, since the characteristic adjustment marks are formed on the measurement-dedicated wafers 11 and 11A to 11C in the above-described embodiments, the exposure mechanism used for manufacturing the characteristic adjustment marks improves the transport mechanism, the wafer holder, and the like. There is no need, and the projection exposure apparatus of FIG. 14 and the overlay error measurement apparatus of FIG. 16 can place the measurement wafer on the XY stage by the wafer loader without providing a special transport mechanism. However, the characteristic adjustment mark may be formed on a plate other than the wafer, and the plate may be attached to and detached from the sample tables 59 and 71 on the XY stage by an operator or the like. Even if the plate on which the characteristic adjustment mark is formed is not a wafer, the shape and size thereof are the same as those of the substrate (wafer or the like) carried into the projection exposure apparatus in FIG. 14 or the overlay error measurement apparatus in FIG. If so, the wafer loader can place the plate on the XY stage. In addition, since an exposure apparatus used for manufacturing a liquid crystal display or the like uses a rectangular substrate, a plate on which a characteristic adjustment mark is formed is not circular but rectangular. Furthermore, the reference plate on which the characteristic adjustment mark is formed may be fixed to a part of the XY stage, and the adjustment may be performed periodically or in accordance with replacement of the diaphragm in the position detection device. . In this case, the time required for the measurement can be shortened as compared with the case where a measurement-dedicated wafer is used.

Further, in the position detection apparatus of FIG. 1, at least one optical element is moved to adjust the optical characteristics of the illumination system and the imaging optical system, but in combination with the movement of the at least one optical element, or A part of the position detection device (optical element) may be replaced with another optical element by itself.
In addition, the exposure apparatus (projection exposure apparatus) of the above embodiment includes an illumination optical system composed of a plurality of lenses, a pedestal that holds the projection optical system by incorporating the projection optical system into the exposure apparatus body and performing optical adjustment. 1 is incorporated into a hardware integrally provided with a wire, and wiring and the like are connected to each other, and optical adjustment is performed as described in each of the above-described embodiments, so that a reticle made up of a large number of mechanical parts. It can be manufactured by attaching a stage or wafer stage to the exposure apparatus main body, connecting wiring and piping, and further performing general adjustment (electrical adjustment, operation check, etc.). The exposure apparatus is preferably manufactured in a clean room where the temperature, cleanliness, etc. are controlled.

  Further, when a semiconductor device is manufactured on a wafer using the exposure apparatus of the above-described embodiment, the semiconductor device includes a step of designing a function / performance of the device, a step of manufacturing a reticle based on this step, silicon A step of producing a wafer from a material, a step of performing alignment with the exposure apparatus of the above-described embodiment and exposing a reticle pattern onto the wafer, a device assembly step (including a dicing process, a bonding process, and a packaging process), an inspection step, etc. It is manufactured through.

  Further, the present invention can be applied not only to the calibration of not only an alignment sensor of a projection exposure apparatus and an overlay error measuring apparatus but also an inspection measuring instrument of various other measurement principles. For example, the present invention can also be applied when adjusting an electron optical system such as a scanning probe microscope using an atomic force microscope (AFM) or the like.

  In addition, this invention is not limited to the above-mentioned embodiment, Of course, a various structure can be taken in the range which does not deviate from the summary of this invention. In addition, Japanese Patent Application No. 10-27474 filed on February 9, 1998 and Japanese Patent Application No. filed on February 16, 1998, each including a description, claims, drawings, and abstract, respectively. The entire disclosures of Japanese Patent Application No. 10-32788 and Japanese Patent Application No. 10-85858 filed on March 31, 1998 are incorporated herein by reference in their entirety.

  According to the measuring method or the adjusting method of the present invention, the mark for characteristic measurement can be easily and accurately formed. Therefore, for example, a position detection device such as an alignment sensor provided in an exposure apparatus used when manufacturing a device such as a semiconductor element can be adjusted with high accuracy using the mark for measuring its characteristics, and overlay accuracy and the like can be improved. In order to improve, devices, such as a highly functional semiconductor device, can be mass-produced with a high yield.

It is the block diagram which made the cross section the part which shows the position detection apparatus of the 1st Embodiment of this invention. It is a top view which shows the several grid-like mark on the wafer used for adjustment in the embodiment. It is a figure which shows the mark for characteristic adjustment of the optical system for a detection in FIG. 2, and the image signal obtained from the image of this mark. It is a figure which shows the mark for characteristic adjustment of the illumination system in FIG. 2, and the image signal obtained from the image of this mark. It is a top view showing a plurality of pairs of test marks on a wafer used for adjustment in a 2nd embodiment of the present invention. FIG. 6 is an enlarged plan view showing two pairs of box-in-box marks 28A and 28B in FIG. It is a figure which shows the mark for characteristic adjustment of the illumination system in FIG. 5, and the image signal obtained from the image of this mark. It is an enlarged plan view which shows the example of another mark which can be used for adjustment in the 2nd Embodiment. FIG. 10 is a plan view showing a plurality of pairs of test marks on another wafer that can be used for adjustment in the second embodiment. It is a figure which shows a pair of mark 29X in FIG. 9, and the image signal obtained from the image of this mark. It is a top view which shows the several mark for evaluation on the wafer used for adjustment in the 3rd Embodiment of this invention. It is a figure which shows the mark for characteristic adjustment of the optical system for a detection in FIG. 11, and the image signal obtained from the image of this mark. It is a figure which shows the mark for characteristic adjustment of the illumination system in FIG. 11, and the image signal obtained from the image of this mark. It is a block diagram which shows the projection exposure apparatus provided with the position detection apparatus of FIG. 1 as an alignment sensor. (A) is a top view which shows an example of the wafer mark used as a detection target with the alignment sensor in FIG. 14, (b) is a top view which shows the other example of the wafer mark used as the detection target with the alignment sensor. It is a block diagram which shows the overlay error measuring apparatus provided with the position detection apparatus of FIG.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Light source, 3 ... (sigma) diaphragm (aperture stop of illumination system), 4a, 4b ... Holding member, 5a, 5b ... (sigma) diaphragm position adjustment mechanism, 10 ... Objective lens group, 11 ... Wafer for adjustment, 13 ... Indicator plate , 16: Frame correction optical system, 16a, 16b: Holding member, 17a, 17b: Frame correction optical system adjustment mechanism, 18: Aperture stop, 22: Imaging element, 23: Control operation system, DM1, DM2: Optical for detection System characteristic adjustment marks, HM1, HM2, ... Illumination system characteristic adjustment marks

Claims (35)

A measurement method for measuring an aberration of a detection optical system of a detection device that detects a mark via a detection optical system,
A test mark in which a first pattern portion including a pattern having a first width and a second pattern portion including a pattern having a second width narrower than the first width are arranged adjacent to each other in the measurement direction. On the other hand, illuminate the illumination light,
Condensing the light beam from the test mark illuminated with the illumination light on a predetermined surface via the detection optical system, and forming an image of the test mark on the predetermined surface,
Based on the image of the test mark formed on the predetermined surface , the interval information between the first pattern portion and the second pattern portion of the test mark is measured,
A measurement method characterized by measuring a predetermined optical characteristic of the detection optical system based on a measurement result of the test mark. The first pattern portion is a lattice pattern including a plurality of line patterns having the first width,
The measurement method according to claim 1, wherein the second pattern portion is a lattice pattern including a plurality of line patterns having the second width.   The measurement method according to claim 1 or 2, wherein the test mark includes the first pattern portion and the second pattern portion that are alternately arranged in the measurement direction.   The measurement method according to claim 3, wherein the test mark is arranged in the measurement direction in the order of the first pattern portion, the second pattern portion, and the first pattern portion.   The measurement method according to claim 1, wherein the test mark has line symmetry with respect to a central axis along a non-measurement direction orthogonal to the measurement direction. The test mark is
A first direction measurement mark in which the first pattern portion and the second pattern portion are alternately arranged in the measurement direction;
The said 1st pattern part and the said 2nd pattern part include the mark for the 2nd direction measurement alternately arrange | positioned in the direction orthogonal to the said measurement direction, The Claim 1 to 5 characterized by the above-mentioned. The measuring method as described in any one. A third pattern portion composed of a pattern in which concave portions and convex portions are alternately arranged periodically in the measurement direction, and a fourth pattern portion composed of a pattern having a shape obtained by inverting the concave portion and the convex portion in the measurement direction. Irradiate illumination light to the second test marks arranged adjacent to each other in parallel,
The light beam from the second test mark illuminated with the illumination light is condensed on a predetermined surface via the detection optical system, and an image of the second test mark is formed on the predetermined surface. ,
Based on the image of the second test mark formed on the predetermined surface, the second test mark is measured,
The characteristic of the illumination system which illuminates the said illumination light with respect to the said 2nd test mark based on the measurement result of the said 2nd test mark is characterized by the above-mentioned. The measuring method as described in.   As the test mark, a mark formed on a measurement-dedicated wafer, a mark formed on a plate detachable from a sample stage, or a mark formed on a reference plate fixed to a part of a stage, It uses, The measuring method as described in any one of Claim 1 to 7 characterized by the above-mentioned.   The measurement method according to claim 1, wherein the test mark is illuminated with broadband illumination light. The predetermined optical property is coma,
An adjustment method comprising adjusting the detection optical system based on the coma aberration measured using the measurement method according to claim 1. The predetermined optical property is asymmetrical aberration;
An adjustment method, comprising: adjusting the detection optical system based on the asymmetric aberration measured using the measurement method according to claim 1.   12. The adjustment method according to claim 10, wherein the detection optical system is adjusted by adjusting a position of a frame correction optical system provided in the detection optical system.   8. An adjustment method, comprising: adjusting a position of a σ stop provided in the illumination system of the detection device based on characteristics of the illumination system measured using the measurement method according to claim 7.   The adjustment method according to claim 10, wherein the detection device is an image processing type alignment sensor.   The adjustment method according to claim 10, wherein the detection device is an image processing type overlay error measurement device. A mark object comprising a test mark for measuring the characteristics of a detection device that detects the mark based on a light beam generated from an illuminated mark,
The test mark is
A first pattern portion including a pattern having a first width;
A second pattern portion including a pattern having a second width narrower than the first width,
The mark object, wherein the first and second pattern portions are arranged adjacent to each other in the measurement direction in order to measure interval information . The first pattern portion is a lattice pattern including a plurality of line patterns having the first width,
The mark object according to claim 16, wherein the second pattern portion is a lattice pattern including a plurality of line patterns having the second width.   The mark object according to claim 16 or 17, wherein the test mark has the first pattern portion and the second pattern portion arranged alternately in the measurement direction.   The mark object according to claim 18, wherein the test mark is arranged in the measurement direction in the order of the first pattern portion, the second pattern portion, and the first pattern portion.   The mark object according to any one of claims 16 to 19, wherein the test mark has line symmetry with respect to a central axis along a non-measurement direction orthogonal to the measurement direction. The test mark is
A first direction measurement mark in which the first pattern portion and the second pattern portion are alternately arranged in the measurement direction;
The first pattern portion and the second pattern portion are provided with second direction measurement marks alternately arranged in a direction orthogonal to the measurement direction. The marked object according to any one of 20.   The mark according to any one of claims 16 to 21, wherein the test mark is a mark for measuring an aberration of a detection optical system of the detection device that receives a light beam generated from the mark. object.   A third pattern portion composed of a pattern in which concave portions and convex portions are alternately arranged alternately in the measurement direction, and a fourth pattern portion composed of a pattern having a shape obtained by inverting the concave portion and the convex portion in the measurement direction. The mark object according to any one of claims 16 to 22, further comprising second test marks arranged adjacent to each other in parallel.   24. The mark object according to any one of claims 16 to 23, wherein the mark object is a measurement-dedicated substrate, a plate detachable from a sample stage, or a reference plate fixed to a part of a stage of a detection apparatus. Marked object described. A detection device for detecting a mark via a detection optical system,
A test mark in which a first pattern portion including a pattern having a first width and a second pattern portion including a pattern having a second width smaller than the first width are arranged adjacent to each other in the measurement direction. A detection apparatus characterized in that imaging is performed through the detection optical system, and predetermined optical characteristics of the detection optical system are adjusted based on the imaging result. The first pattern portion is a lattice pattern including a plurality of line patterns having the first width,
26. The detection device according to claim 25, wherein the second pattern unit is a lattice pattern including a plurality of line patterns having the second width.   27. The detection device according to claim 25, wherein the test mark includes the first pattern portion and the second pattern portion arranged alternately in the measurement direction.   The detection apparatus according to claim 27, wherein the test mark is arranged in the measurement direction in the order of the first pattern portion, the second pattern portion, and the first pattern portion.   The detection device according to any one of claims 25 to 28, wherein the test mark has line symmetry with respect to a central axis along a non-measurement direction orthogonal to the measurement direction. The test mark is
A first direction measurement mark in which the first pattern portion and the second pattern portion are alternately arranged in the measurement direction;
The said 1st pattern part and the said 2nd pattern part are equipped with the mark for the 2nd direction measurement arrange | positioned alternately in the direction orthogonal to the said measurement direction, From Claim 25 characterized by the above-mentioned. 30. The detection device according to any one of 29.   31. The detection apparatus according to claim 25, wherein the predetermined optical characteristic is an aberration of the detection optical system.   32. The detection device according to claim 25, wherein the detection device is an image processing type alignment sensor.   32. The detection apparatus according to claim 25, wherein the detection apparatus is an image processing type overlay error measurement apparatus.   The detection device according to any one of claims 25 to 33, wherein the detection device includes a reference plate on which the test mark is formed.   The detection mark according to any one of claims 25 to 33, wherein the test mark is formed on a measurement-dedicated wafer that can be measured by the detection device, or a plate that can be attached to and detached from a sample stage. apparatus.

Images