Title:

Mark position detecting system and method for detecting mark position

Document Type and Number:

United States Patent Application 20010026368

Kind Code:

Abstract:

A mark position detecting system comprises a light source for illuminating light L1 on a mark for alignment measurement formed on a semiconductor substrate, a light detecting optical system for detecting reflected light from the mark and a computer including a measured waveform preparing part, theoretical waveform preparing part, waveform comparison part and misalignment operation part. The measured waveform preparation part prepares and output a measured waveform denoting reflected light strength distribution dependent on a cross sectional shape of the mark on a basis of a detection result by the detecting optical system. The theoretical waveform preparing part prepares a theoretical waveform on the light strength distribution of the reflected light from a region of the mark where the strength thereof would change, on the basis of information on the material of SiO2 layer of the mark and the thickness T1 thereof, the material of SiN layer of the mark, the thickness T2 thereof and the depth D1, D2 of the concavity C1, C2. The waveform comparison part compares the theoretical waveform with the measured waveform on the basis of positional information on a region where the light strength of the reflected light changes. The misalignment calculating part calculates a misalignment quantity between the theoretical waveform and the measured waveform. The quantity of the misalignment between the theoretical waveform and the measured waveform turns out to be zero whether the cross sectional shape of the alignment measurement mark is symmetric or non-symmetric. Thus the position of the mark is accurately detected.

Representative Image:

Mark position detecting system and method for detecting mark position

Inventors:

Mikami, Toru (Fujisawa-Shi, JP)

Plaque It!

What is claimed is:

1. A mark position detecting system comprising: a light emitter for emitting light to irradiate a mark for misalignment measurement, the mark being formed on a semiconductor substrate, shape information of the mark and material information of an element constituting the mark are previously given; a light detector for detecting a reflected beam of light emitted from the mark on irradiation of the light; a waveform recognition part for preparing a measured waveform on the basis of the detected result of said light detector, the measured waveform denoting strength distribution of said reflected beam according to the shape and the material of the mark; a theoretical waveform preparing part for preparing a theoretical reflected beam waveform on the basis of said shape information and said material information of the mark, said theoretical reflected beam waveform denoting theoretical strength distribution of said reflected beam which would be obtained by irradiating a desired region of the mark with the light; and a determining part for comparing said measured waveform with said theoretical reflected beam waveform to acquire positional information on a place on a surface of the substrate, the place corresponding to the portion of said measured waveform which is most similar to said theoretical reflected beam waveform and for detecting the position of the mark on the basis of the acquired said positional information.

2. A mark position detecting system according to claim 1, wherein said theoretical waveform preparing part prepares said theoretical reflected beam waveform of a spot at which strength of said reflected beam changes.

3. A mark position detecting system according to claim 2, wherein; the mark includes a first thin film formed of a first material on the substrate and a second thin film formed of a second material on the first film so as to protrude from the first film, the first thin film being provided thereon with a first concavity having a first depth and a second concavity having a second depth, the first concavity and the second concavity are spaced from each other, and, the second thin film being arranged in the middle of the first and second concavities, said shape information includes step information concerning a thickness of the first thin film, a thickness of the second thin film, the first depth and the second depth, said theoretical waveform preparing part prepares a first through a fourth theoretical reflected beam waveforms, the first theoretical reflected beam waveform corresponding to a first place which equivalent to an outside edge of the first concavity in view of the second thin film, the second theoretical reflected beam waveform corresponding to a second place which equivalent to an outside edge of the second concavity in view of the second thin film, the third theoretical reflected beam waveform corresponding to a third place equivalent to a first sidewall of the second thin film and the fourth theoretical reflected beam waveform corresponding to a fourth place equivalent to a second sidewall of the second thin film, the second sidewall being faced to the first sidewall, and said determining part compares the measured waveform with the first through fourth theoretical reflected beam waveforms respectively, calculates a first middle point position which is the middle point of a first edge position corresponding to the first place and a second edge position corresponding to the second place, calculates a second middle point which is the middle point of the first sidewall position corresponding to the third place and the second sidewall position corresponding to the fourth place, and determines whether any alignment occurs between said first middle point and said second middle point.

4. A mark position detecting system comprising: a light emitter for emitting light to irradiate a mark for misalignment measurement, the mark being formed on a semiconductor substrate, material information of an element constituting a surface portion of the mark being previously given; a spectroscope for diffracting a reflected beam of light into a ray having an arbitrary wavelength, said reflected beam being emitted from the mark on irradiation of the light; a first light detector for detecting the diffracted ray diffracted by said spectroscope; a shape information acquiring part for receiving the detected result of said first light detector and said material information, recognizing a measured diffracted ray waveform denoting strength distribution of said diffracted ray according to the shape and the material of the mark and for acquiring shape information of the mark by analyzing said measured diffracted ray waveform; a second light detector for detecting said reflected beam, said reflected beam being light emitted from said light emitter and reflected on the mark; a waveform recognition part for preparing a measured waveform on the basis of the detected result of said second light detector, the measured waveform denoting strength distribution of said reflected beam according to the shape and the material of the mark; a theoretical waveform preparing part for preparing a theoretical diffracted ray waveform which is a theoretical waveform of said diffracted ray on the basis of said material information, for supplying said theoretical diffracted ray waveform to said shape information acquiring part and for preparing a theoretical reflected beam waveform on the basis of said shape information given from said shape information acquiring part and said material information, said theoretical reflected beam waveform denoting theoretical strength distribution of said reflected beam which would be obtained by irradiating a desired region of the mark with the light; and a determining part for comparing said measured waveform with said theoretical reflected beam waveform to acquire positional information on a place on a surface of the substrate, the place corresponding to the portion of said measured waveform which is most similar to said theoretical reflected beam waveform, and for detecting the position of the mark on the basis of the acquired said positional information.

5. A mark position detecting system according to claim 4, wherein said theoretical waveform preparing part prepares said theoretical reflected beam waveform of a spot at which strength of said reflected beam changes.

6. A mark position detecting system according to claim 5, which further comprises a parameter calculating part for generating a plurality of parameters capable of being candidates to said shape information and for supplying said parameters to said shape information acquiring part, wherein said theoretical waveform preparing part prepares said theoretical diffracted ray waveform on the basis of said material information every said parameter, and said shape information acquiring part compares said measured diffracted ray waveform with said theoretical diffracted ray waveform of every said parameter, selects said theoretical diffracted ray waveform which is most similar to said measured diffracted ray waveform of said theoretical diffracted ray waveforms and determines the parameter of said selected theoretical diffracted ray waveform as said shape information.

7. A mark position detecting system according to claim 6, wherein; the mark includes a first thin film formed of a first material on the substrate and a second thin film formed of a second material on the first film so as to protrude from the first film, the first thin film being provided thereon with a first concavity having a first depth and a second concavity having a second depth, the first concavity and the second concavity are spaced from each other, and, the second thin film being arranged in the middle of the first and second concavities, said shape information includes step information concerning a thickness of the first thin film, a thickness of the second thin film, the first depth and the second depth, said theoretical waveform preparing part prepares a first through a fourth theoretical reflected beam waveforms, the first theoretical reflected beam waveform corresponding to a first place which equivalent to an outside edge of the first concavity in view of the second thin film, the second theoretical reflected beam waveform corresponding to a second place which equivalent to an outside edge of the second concavity in view of the second thin film, the third theoretical reflected beam waveform corresponding to a third place equivalent to a first sidewall of the second thin film and the fourth theoretical reflected beam waveform corresponding to a fourth place equivalent to a second sidewall of the second thin film, the second sidewall being faced to the first sidewall, and said determining part compares the measured waveform with the first through fourth theoretical reflected beam waveforms respectively, calculates a first middle point position which is the middle point of a first edge position corresponding to the first place and a second edge position corresponding to the second place, calculates a second middle point which is the middle point of the first sidewall position corresponding to the third place and the second sidewall position corresponding to the fourth place, and determines whether any alignment occurs between said first middle point and said second middle point.

8. A method of detecting a mark position, the mark being formed on a semiconductor substrate for misalignment measurement, said method comprising steps: acquiring material information on an element constituting the mark; acquiring shape information on the mark; irradiating the mark with light; detecting a reflected beam of light emitted from the mark on irradiation of the light; acquiring a measured waveform denoting strength distribution of said reflected beam according to the shape and the material of the mark on the basis of the detected result of said reflected beam; preparing a theoretical reflected beam waveform on the basis of said shape information and said material information on the mark, said theoretical reflected beam waveform denoting theoretical strength distribution of said reflected beam which would be obtained by irradiating a desired region of the mark with the light; comparing said measured waveform with said theoretical reflected beam waveform to acquire positional information on a place on a surface of the substrate, the place corresponding to the portion of said measured waveform which is most similar to said theoretical reflected beam waveform; and detecting the position of the mark on the basis of the acquired said positional information.

9. A method of detecting a mark position according to claim 8, wherein said desired region of the mark is a region where strength of said reflected beam changes.

10. A method of detecting a mark position according to claim 9, wherein said step of acquiring shape information on the mark includes steps; diffracting said reflected beam into a ray having an arbitrary wavelength and detecting said diffracted ray in accordance with a surface shape and a material of the mark, recognizing a measured diffracted ray waveform denoting strength distribution of said diffracted ray and acquiring said shape information on the mark by analyzing said measured diffracted ray waveform.

11. A method of detecting a mark position according to claim 10, wherein said step of acquiring said shape information includes steps of; generating a plurality of parameters capable of being candidates to said shape information, preparing a theoretical diffracted ray waveform on the basis of said material information every said parameter, said theoretical diffracted ray waveform being a theoretical waveform of said diffracted ray, comparing said measured diffracted ray waveform with said theoretical diffracted ray waveform of every said parameter, selecting said theoretical diffracted ray waveform which is most similar to said measured diffracted ray waveform of said theoretical diffracted ray waveforms, and determining the parameter of said selected theoretical diffracted ray waveform as said shape information.

12. A method of detecting a mark position according to claim 11, wherein; the mark includes a first thin film formed of a first material on the substrate and a second thin film formed of a second material on the first film so as to protrude from the first film, the first thin film being provided thereon with a first concavity having a first depth and a second concavity having a second depth, the first concavity and the second concavity are spaced from each other, and, the second thin film being arranged in the middle of the first and second concavities, said shape information includes step information concerning a thickness of the first thin film, a thickness of the second thin film, the first depth and the second depth, said step of preparing theoretical reflected beam waveform is a step of preparing a first through a fourth theoretical reflected beam waveforms, the first theoretical reflected beam waveform corresponding to a first place which equivalent to an outside edge of the first concavity in view of the second thin film, the second theoretical reflected beam waveform corresponding to a second place which equivalent to an outside edge of the second concavity in view of the second thin film, the third theoretical reflected beam waveform corresponding to a third place equivalent to a first sidewall of the second thin film and the fourth theoretical reflected beam waveform corresponding to a fourth place equivalent to a second sidewall of the second thin film, the second sidewall being faced to the first sidewall, and said step of detecting the position of the mark is a step of calculating a first middle point position which is the middle point of a first edge position corresponding to the first place and a second edge position corresponding to the second place, calculating a second middle point which is the middle point of the first sidewall position corresponding to the third place and the second sidewall position corresponding to the fourth place, and determines whether any alignment occurs between said first middle point and said second middle point.

Description:

CROSS REFERENCE TO RELATED APPLICATION

[0001] This application claims benefit of priority under 35USC §119 to Japanese patent application No. 2000-086908, filed on Mar. 27, 2000, the contents of which are incorporated by reference herein.

BACKGROUND OF THE INVENTION

[0002] 1. Field of the Invention

[0003] The present invention relates to a system and a method for detecting misalignment between masks in manufacturing a semiconductor device. In more specific, the present invention relates to a system and a method for detecting a position of a misalignment measurement mark which is previously formed on a semiconductor substrate.

[0004] 2. Description of the Prior Art

[0005] In manufacturing a semiconductor device with lamination layers stacked with thin films having different patterns respectively, it is greatly important to accurately align a mask (reticle) with a semiconductor substrate for forming a pattern thereon.

[0006] For such mask alignment, a commonly adapted method has steps of previously forming a misalignment measurement mark (which will hereinafter be occasionally referred as a measurement mark) in a region other than that for forming a device on a semiconductor substrate, detecting the position of the measurement mark and adjusting a mask alignment position on the basis of the detected position.

[0007] Conventional methods for detecting a position of an alignment measurement mark will hereinafter be described taking a slice level method and a correlation method for instances. In the following respective figures, the same reference numbers are given to the same portions, and the descriptions thereof are appropriately omitted.

[0008] FIG. 1 is a schematic diagram showing a conventional measurement mark position detecting system. The misalignment measurement mark position detecting system 110 shown in this figure comprises; a light source 13 , a half mirror 15 , a stage 70 , a CCD (Charge Coupled Device) sensor 33 , an A-D (Analogue to Digital) converter 35 and a control computer 110 . A Si (silicon) substrate 120 is supported on the stage 70 . The substrate 120 is previously provided with a measurement mark 20 which is an object to be measured. In this figure a cross section diagram of the measurement mark 20 is shown taken along a line in the X direction and portions of the substrate 120 other than the measurement mark 20 are omitted.

[0009] FIG. 2 is an enlarged view of the measurement mark 20 shown in FIG. 1 . As shown in FIG. 2 , the measurement mark 20 includes a SiO 2 layer 23 formed on the Si substrate 120 , and a SiN layer 27 which on the SiO 2 layer 23 so as to protrude therefrom. The SiO 2 layer 23 and the SiN layer 27 are formed in thickness of T 1 and T 2 respectively, and each values thereof are 1 μm in this example. Two concavities C 1 and C 2 are formed on a surface of the SiO 2 layer 23 . These concavities have depth D 1 and D 2 of 0.12 μm respectively and thus constitute steps. The SiN layer 27 is arranged such that the center thereof is positioned right in the middle of concavities C 1 and C 2 in the cross section view of the FIG. 2 . That is, from the point of view of the SiN layer 27 the SiN layer 27 is arranged such that the center thereof is positioned right in the middle of outside edges E 1 , E 4 of the concavities. The measurement mark 20 thus forms a symmetry shape with respect to the centerline 11 of SiN layer 27 .

[0010] The position of the measurement mark can be detected by detecting the center point P 1 on the top surface of the SiN layer 27 . However, a typical method of detecting the point P 1 includes a step of recognizing that the center point P 1 of the SiN layer 27 coincides with the middle point of the outside edges E 1 , E 4 of concavities C 1 , C 2 .

[0011] (1) Slice Level Method

[0012] Referring FIGS. 3B, 4 and 6 showing waveforms and the flow-chart of FIG. 5, a slice level method for detecting a measurement mark will be described.

[0013] First, using the system 100 , a beam of light L 1 having a predetermined wavelength λ or white light is emitted from the light source 13 to irradiate the measurement mark 20 via the half-mirror 15 (step S 101 ). A reflected beam of light L 2 is generated from the measurement mark 20 . The reflected beam L 2 passes through the half-mirror 15 and is detected by a CCD sensor 33 (step S 102 ). The reflected beam L 2 includes a ray from the interface between Si substrate 120 and SiO 2 layer 23 , a ray from the surface of the SiO 2 layer 23 , a ray from the interface between SiO 2 layer 23 and SiN layer 27 and a ray from the surface of the SiN layer 27 . Since above mentioned rays interfere each other, the reflected beam L 2 enters the CCD sensor 33 as the beam having various light strength dependent on each difference between the optical path lengths from these interfaces or surfaces to a pixel portion of the CCD sensor 33 .

[0014] In the CCD sensor 33 pixels are arranged in a row in the x direction. Electric charges are generated from each pixel in response to the rays of the reflected beam entering the pixel. Signals from these charges are conveyed to the control computer 110 through the A/D converter 35 .

[0015] The control computer 110 processes the signals supplied from the CCD sensor 33 to recognize a waveform in a diagram with a horizontal axis and a vertical axis. The horizontal axis denotes X coordinates of the measurement mark in the X direction and the vertical axis denotes strengths of the reflected beam from the measurement mark (step S 103 ). A position coordinate of the measurement mark with respect to the substrate 120 (which will hereinafter referred to as a wafer position coordinate) is detected in a conventional way.

[0016] FIG. 3B shows a waveform diagram obtained by the control computer 110 together with the shape of the measurement mark in a cross-sectional view. As shown in FIG. 3 B, each position coordinate on the horizontal axis corresponds to a positional coordinate of the measurement mark respectively. For example, edges E 1 through E 6 of the convexo-concave shape correspond to X 1 through X 6 of the waveform figure respectively.

[0017] As shown in FIG. 3 B, assuming that the light strength of the reflected beam from the concavity C 1 corresponding the position coordinates from X 1 to X 2 is rd 1 , and that the light strength of the reflected beam from the concavity C 2 corresponding to the position coordinates form X 5 to X 6 is rd 2 , and that the light strength of the reflected beam from the other surface of the SiO 2 layer 23 is r 0 , the following correlation exists between these strengths.

ro>rd 1 , rd 2 (1)

rd 1 =rd 2 (2)

[0018] Thus, the waveform of the reflected beam obtained from the measurement mark having a line symmetry shape in a cross section view has a concavity portion in shape in and near the region of the position coordinates from X 1 to X 2 and a concavity portion in and near the region of the position coordinates from X 5 to X 6 . The entire waveform has a line symmetry shape along a line 11 ′ which passes the middle point X 34 of X 3 and X 4 and is perpendicular to the X-axis.

[0019] Referring now to FIG. 4 and FIG. 5, a method for processing a waveform in such a symmetry shape and for detecting the position of the measurement mark 20 by means of a slice level method will be described below.

[0020] First, the position XM 1 where the light strength drops most sharply in and near a region having position coordinates from X 1 to X 2 in the waveform figure is detected (step S 104 ).

[0021] Similarly, the position XM 6 where the light strength rises most sharply in and near a region having position coordinates from X 5 to X 6 in the waveform figure is detected (step S 105 ).

[0022] Next, the middle position XM 16 of the position XM 1 and the position XM 6 acquired at above-mentioned step is calculated (step S 106 ).

[0023] Then, at steps similar to the above steps S 104 through S 0106 , the position XM 3 where the light strength drops most sharply in a portion having position coordinate of and near the X 3 , and the position XM 4 where the light strength rises most sharply in a portion having position coordinate of and near X 4 are detected respectively (steps S 107 and S 108 ). Then a middle position XM 34 of the X 3 and the X 4 is calculated (step S 109 ).

[0024] At last, the difference between XM 34 and XM 16 is calculated and the calculated value is outputted as misalignment (step S 110 ).

[0025] In the example shown in the waveform diagram of FIG. 4 , since the waveform of the light strength of the reflected beam has a symmetric shape, XM 1 and the position coordinate X 1 , and XM 6 and the position coordinate X 6 coincide with each other respectively. Therefore, it is recognized that XM 34 −XM 16 =0 and that MX 16 accurately coincides with the middle point of X 1 and X 6 . As a result, the position of the measurement mark can accurately be detected, so that a mask can precisely be aligned with a substrate or a pattern previously formed thereon in a subsequent lithography process.

[0026] (2) Correlation Method

[0027] Next, a correlation method for detecting a measurement mark will be described referring to FIGS. 1, 3B and 6 showing waveforms and a flow-chart of FIG. 6 .

[0028] Steps S 111 through S 113 of obtaining a waveform figure shown in FIG. 3B by irradiating the measurement mark with the light L 1 and by detecting the reflected beam L 2 with the CCD sensor 33 are substantially the same as steps S 101 through S 103 , and each of step numbers of steps S 111 through S 113 is that added by 5 to each corresponding step shown in FIG. 5 .

[0029] As shown in FIG. 6, a portion 11 of the waveform having the position coordinates of and near X 1 is extracted. Then, the waveform portion 11 is reversed with respect to a line intersecting X 1 and perpendicular to the X-axis by means of a mirror-reversing process to prepare a symmetric graphic. Data on the prepared graphic are then stored in a memory (not shown) as a reference waveform 11 inv (step S 114 ).

[0030] Next, a portion of the waveform having the position coordinates X 5 to X 6 and position coordinates in the vicinity hereof is compared with the reference waveform 11 inv and the waveform which is most similarity to the reference waveform is detected. Then the position coordinate correspondent to the detected waveform is designated as XN 6 (step S 115 ).

[0031] Then, a middle point of XN 1 corresponding to the waveform portion 11 and the position coordinate obtained at step S 115 is calculated and is designated as the middle point XN 16 of the position XN 1 and the position XN 6 (step S 116 ).

[0032] Then, the middle point XN 34 of the position X 3 and the position X 4 is calculated at steps similar to the above-mentioned steps S 114 through S 116 (steps S 117 through S 119 ).

[0033] At last, the difference between XN 34 and XN 16 is calculated and the calculated value is outputted as a quantity of misalignment (step S 120 ).

[0034] By means of the correlation method described above, when a waveform obtained from the reflected beam is symmetric, XN 16 coincides exactly with the middle point XN 34 of the position coordinate X 3 and the position coordinate X 4 , so that it is possible to accurately detect the position of the measurement mark 20 .

[0035] However, both the slice level method and the correlation method which are described above have a problem that a mark position cannot accurately be detected when a measurement mark has a non-symmetric shape. This problem will be described in more detail below.

[0036] FIG. 7 shows an example of a misalignment measurement mark having a non-symmetric cross sectional shape. Materials and film thickness of elements constituting the measurement mark 21 shown in FIG. 7 are the same as those of the measurement mark 20 shown in FIG. 2 . And the fact that SiN layer 27 is arranged right in the middle of two concavities C 3 and C 4 is also the same as the aforementioned measurement mark 21 .

[0037] However, depths D 1 ′ and D 2 ′ of the two concavities C 3 and C 4 arranged on SiO 2 layer 24 of the measurement mark 21 are different from those of the measurement mark 20 . In specific, the concavities C 3 and C 4 are formed in depths D 1 ′=0.1 μm and D 2 ′=0.14 μm respectively. Due to such constitution the measurement mark 21 has a non-symmetric cross sectional shape with respect to the centre line 12 of SiN layer 27 . For this reason, when strength distribution of the reflected beam from the measurement mark 21 is obtained at steps S 101 through S 103 shown in FIGS. 5 and 11 , the waveform thereof is then acquired as shown in FIG. 8B because a phase of the reflected beam from the portion of the concavity C 4 reverses.

[0038] When the position of the measurement mark 21 is intended to be detected using the waveform shown in FIG. 8B by means of conventional methods, following problems occur.

Slice Level Method

[0039] As shown in FIG. 9 , XM 1 corresponds to the position coordinate of X 1 at the step of detecting the position coordinate XM 1 where the strength of the reflected beam drops most sharply in and near the position coordinates from X 1 to X 2 (step S 104 in FIG. 5 ).

[0040] However, XM 6 corresponds not to the position coordinate X 6 but to the position coordinate X 5 at a step of detecting a position coordinate XM 6 where the rise of the light strength is expected to be most steep in and near the position coordinates from X 5 to X 6 (step S 105 in FIG. 5 ). The middle point thereof then corresponds not to the middle position of the position coordinate X 1 and X 6 but to the middle position of the position coordinate X 1 and X 5 . For this reason, XM 16 never coincides with the middle point XM 34 of XM 3 and MX 4 which are obtained at steps S 107 through S 109 , and an error occurs by a distance of XE shown in FIG. 9 . As a result, this error renders it impossible to accurately detect a misalignment with a mask in a subsequent process.

Correlation Method

[0041] As can be seen from FIG. 10, a mirror-reversed waveform 13 inv ′ (not shown) of a waveform portion having position coordinates of and near X 3 is most similar to a portion of the waveform having position coordinates of and near X 4 . The middle position XN 34 of the detected position coordinates corresponds to the middle point of X 3 and X 4 similarly to the example of symmetric cross sectional shape.

[0042] However, a mirror-reversed waveform 11 inv ′ prepared from a portion of the waveform 11 ′ having position coordinates of and near X 1 is most similar to that having position coordinates of and near X 5 . Then the middle point XN 16 thereof corresponds not to the middle point of the position coordinate X 1 and X 6 but to the middle point of the position coordinate X 1 and X 5 . For this reason, as shown in FIG. 10 , an error occurs by a distance of XE between XN 16 and the middle pointXN 34 of XN 3 and MN 4 obtained at steps S 117 through S 119 shown in FIG. 11 . As a result, this error renders it impossible to precisely detect misalignment between the measurement mark 21 and a mask.

[0043] As mentioned above, according to the conventional methods, a position of a measurement mark can accurately detected when a cross sectional shape of the mark is symmetric, however, there is a problem that a position of a mark having a non-symmetric cross sectional can not be precisely detected.

SUMMARY OF THE INVENTION

[0044] It is therefore an object of the present invention to provide a mark position detecting system which can accurately detect the position of an alignment measurement mark even if the cross sectional shape thereof is not symmetric.

[0045] It is another object of the present invention to provide a method for precisely detecting the position of an alignment measurement mark even if the cross sectional shape thereof is not symmetric.

[0046] According to a first aspect of the present invention, there is provided a mark position detecting system comprising: a light emitter for emitting light to irradiate a mark for misalignment measurement, the mark being formed on a semiconductor substrate, shape information of the mark and material information of an element constituting the mark are previously given; a light detector for detecting a reflected beam of light emitted from the mark on irradiation of the light; a waveform recognition part for preparing a measured waveform on the basis of the detected result of the light detector, the measured waveform denoting strength distribution of the reflected beam according to the shape and the material of the mark; a theoretical waveform preparing part for preparing a theoretical reflected beam waveform on the basis of the shape information and the material information of the mark, the theoretical reflected beam waveform denoting theoretical strength distribution of the reflected beam which would be obtained by irradiating a desired region of the mark with the light; and a determining part for comparing the measured waveform with the theoretical reflected beam waveform to acquire positional information on a place on a surface of the substrate, the place corresponding to the portion of the measured waveform which is most similar to the theoretical reflected beam waveform and for detecting the position of the mark on the basis of the acquired the positional information.

[0047] Because the theoretical waveform preparing part prepares the theoretical reflected beam waveform on the basis of the shape information and the material information of the mark, and the determining part compares the measured waveform with the theoretical reflected beam waveform, it is possible to accurately acquire a positional information on a desired place on the substrate for specifying the mark. Therefore, the position of the mark can be precisely detected whether a cross sectional shape of the mark is, for example symmetric or non-symmetric.

[0048] The theoretical waveform preparing part may preferably prepare the theoretical reflected beam waveform of a spot at which strength of the reflected beam changes. This enables to obtain positional information on a characteristic place of the mark.

[0049] In a preferred embodiment of the present invention, the mark includes a first thin film formed of a first material on the substrate and a second thin film formed of a second material on the first film so as to protrude from the first film, the first thin film being provided thereon with a first concavity having a first depth and a second concavity having a second depth, the first concavity and the second concavity are spaced from each other, and, the second thin film being arranged in the middle of the first and second concavities, the shape information includes step information concerning a thickness of the first thin film, a thickness of the second thin film, the first depth and the second depth, the theoretical waveform preparing part prepares a first through a fourth theoretical reflected beam waveforms, the first theoretical reflected beam waveform corresponding to a first place which equivalent to an outside edge of the first concavity in view of the second thin film, the second theoretical reflected beam waveform corresponding to a second place which equivalent to an outside edge of the second concavity in view of the second thin film, the third theoretical reflected beam waveform corresponding to a third place equivalent to a first sidewall of the second thin film and the fourth theoretical reflected beam waveform corresponding to a fourth place equivalent to a second sidewall of the second thin film, the second sidewall being faced to the first sidewall, and the determining part compares the measured waveform with the first through fourth theoretical reflected beam waveforms respectively, calculates a first middle point position which is the middle point of a first edge position corresponding to the first place and a second edge position corresponding to the second place, calculates a second middle point which is the middle point of the first sidewall position corresponding to the third place and the second sidewall position corresponding to the fourth place, and determines whether any alignment occurs between the first middle point and the second middle point.

[0050] When a mark in the above mentioned shape is used and the determining part determines whether any alignment occurs between the first middle point and the second middle point, it is possible to confirm whether there is any detected error or not, so that the position of the mark can be detected with a high degree of accuracy.

[0051] According to a second aspect of the present invention, there is provided a mark position detecting system comprising: a light emitter for emitting light to irradiate a mark for misalignment measurement, the mark being formed on a semiconductor substrate, material information of an element constituting a surface portion of the mark being previously given; a spectroscope for diffracting a reflected beam of light into a ray having an arbitrary wavelength, the reflected beam being emitted from the mark on irradiation of the light; a first light detector for detecting the diffracted ray diffracted by the spectroscope; a shape information acquiring part for receiving the detected result of the first light detector and the material information, recognizing a measured diffracted ray waveform denoting strength distribution of the diffracted ray according to the shape and the material of the mark and for acquiring shape information of the mark by analyzing the measured diffracted ray waveform; a second light detector for detecting the reflected beam, the reflected beam being light emitted from the light emitter and reflected on the mark; a waveform recognition part for preparing a measured waveform on the basis of the detected result of the second light detector, the measured waveform denoting strength distribution of the reflected beam according to the shape and the material of the mark; a theoretical waveform preparing part for preparing a theoretical diffracted ray waveform which is a theoretical waveform of the diffracted ray on the basis of the material information, for supplying the theoretical diffracted ray waveform to the shape information acquiring part and for preparing a theoretical reflected beam waveform on the basis of the shape information given from the shape information acquiring part and the material information, the theoretical reflected beam waveform denoting theoretical strength distribution of the reflected beam which would be obtained by irradiating a desired region of the mark with the light; and a determining part for comparing the measured waveform with the theoretical reflected beam waveform to acquire positional information on a place on a surface of the substrate, the place corresponding to the portion of the measured waveform which is most similar to the theoretical reflected beam waveform, and for detecting the position of the mark on the basis of the acquired the positional information.

[0052] According to the second aspect, the mark position detecting system further comprises the shape information acquiring part, so that the shape information of a misalignment mark can also be acquired with a single system. Therefore, it is possible to detect a position of the mark with high throughput.

[0053] It is advantageous that the mark position detecting system in the second aspect of the invention further comprises a parameter calculating part for generating a plurality of parameters capable of being candidates to the shape information and for supplying the parameters to the shape information acquiring part, wherein the theoretical waveform preparing part prepares the theoretical diffracted ray waveform on the basis of the material information every the parameter, and the shape information acquiring part compares the measured diffracted ray waveform with the theoretical diffracted ray waveform of every the parameter, selects the theoretical diffracted ray waveform which is most similar to the measured diffracted ray waveform of the theoretical diffracted ray waveforms and determines the parameter of the selected theoretical diffracted ray waveform as the shape information.

[0054] According to a third aspect of the present invention, there is provided a method of detecting a mark position, the mark being formed on a semiconductor substrate for misalignment measurement, the method comprising steps: acquiring material information on an element constituting the mark; acquiring shape information on the mark; irradiating the mark with light; detecting a reflected beam of light emitted from the mark on irradiation of the light; acquiring a measured waveform denoting strength distribution of the reflected beam according to the shape and the material of the mark on the basis of the detected result of the reflected beam; preparing a theoretical reflected beam waveform on the basis of the shape information and the material information on the mark, the theoretical reflected beam waveform denoting theoretical strength distribution of the reflected beam which would be obtained by irradiating a desired region of the mark with the light; comparing the measured waveform with the theoretical reflected beam waveform to acquire positional information on a place on a surface of the substrate, the place corresponding to the portion of the measured waveform which is most similar to the theoretical reflected beam waveform; and detecting the position of the mark on the basis of the acquired the positional information.

[0055] According to the third aspect of the invention, a theoretical reflected beam waveform is prepared the basis of the shape information and the material information on the mark, so that it is possible to accurately acquire a positional information on a desired place on the substrate for specifying the mark. Therefore, the position of the mark can be precisely detected whether a cross sectional shape of the mark is, for example symmetric or non-symmetric.

[0056] In the mark position detecting method the step of acquiring shape information on the mark may preferably include steps;

[0057] diffracting the reflected beam into a ray having an arbitrary wavelength and detecting the diffracted ray in accordance with a surface shape and a material of the mark, recognizing a measured diffracted ray waveform denoting strength distribution of the diffracted ray and acquiring the shape information on the mark by analyzing the measured diffracted ray waveform.

[0058] Thus, the shape information of the misalignment mark can also be acquired in a series of steps, it is possible to detect a position of the mark with high throughput.

[0059] In a preferred embodiment of the method of detecting a mark position, the step of acquiring the shape information includes steps of; generating a plurality of parameters capable of being candidates to the shape information, preparing a theoretical diffracted ray waveform on the basis of the material information every the parameter, the theoretical diffracted ray waveform being a theoretical waveform of the diffracted ray, comparing the measured diffracted ray waveform with the theoretical diffracted ray waveform of every the parameter, selecting the theoretical diffracted ray waveform which is most similar to the measured diffracted ray waveform of the theoretical diffracted ray waveforms, and determining the parameter of the selected theoretical diffracted ray waveform as the shape information.

[0060] Furthermore, in a further preferred embodiment of the method of detecting a mark position, the mark includes a first thin film formed of a first material on the substrate and a second thin film formed of a second material on the first film so as to protrude from the first film, the first thin film being provided thereon with a first concavity having a first depth and a second concavity having a second depth, the first concavity and the second concavity are spaced from each other, and, the second thin film being arranged in the middle of the first and second concavities, the shape information includes step information concerning a thickness of the first thin film, a thickness of the second thin film, the first depth and the second depth, the step of preparing theoretical reflected beam waveform is a step of preparing a first through a fourth theoretical reflected beam waveforms, the first theoretical reflected beam waveform corresponding to a first place which equivalent to an outside edge of the first concavity in view of the second thin film, the second theoretical reflected beam waveform corresponding to a second place which equivalent to an outside edge of the second concavity in view of the second thin film, the third theoretical reflected beam waveform corresponding to a third place equivalent to a first sidewall of the second thin film and the fourth theoretical reflected beam waveform corresponding to a fourth place equivalent to a second sidewall of the second thin film, the second sidewall being faced to the first sidewall, and

[0061] the step of detecting the position of the mark is a step of calculating a first middle point position which is the middle point of a first edge position corresponding to the first place and a second edge position corresponding to the second place, calculating a second middle point which is the middle point of the first sidewall position corresponding to the third place and the second sidewall position corresponding to the fourth place, and determines whether any alignment occurs between the first middle point and the second middle point.

[0062] When a mark in the above mentioned shape is used and it is determined whether any alignment occurs between the first middle point and the second middle point, it is possible to confirm whether there is any detected error or not, so that the position of the mark can be detected with a high degree of accuracy.

BRIEF DESCRIPTION OF THE DRAWINGS

[0063] The present invention will be understood more fully from the detailed description given below and from the accompanying drawings of the preferred embodiments of the invention. However, the drawings are not intended to imply limitation of the invention to a specific embodiment, but are for explanation and understanding only.

[0064] In the drawings:

[0065] FIG. 1 is a schematic diagram showing a conventional measurement mark position detecting system;

[0066] FIG. 2 is an enlarged view of the measurement mark shown in FIG. 1 ;

[0067] FIG. 3A is a cross-sectional view showing the shape of the measurement mark shown in FIG. 2 and FIG. 3B shows an example of a waveform denoting light strength of reflected beam from the measurement mark shown in FIG. 3A in association with position coordinates of the measurement mark;

[0068] FIG. 4 is a schematic diagram showing a waveform for explaining the slice level method;

[0069] FIG. 5 is a flow chart for explaining the slice level method;

[0070] FIG. 6 is a schematic diagram showing a waveform for explaining the correlation method;

[0071] FIG. 7 is an illustration showing an example of a measurement mark having a non-symmetric cross sectional shape;

[0072] FIG. 8A is a cross-sectional view showing the shape of the measurement mark shown in FIG. 7 and FIG. 8B is a schematic diagram showing a waveform denoting light strength of reflected light from the measurement mark shown in FIG. 7 in association with position coordinates of the measurement mark;

[0073] FIG. 9 is a schematic diagram showing a waveform for explaining the slice level method;

[0074] FIG. 10 is a schematic diagram showing a waveform for explaining the correlation method;

[0075] FIG. 11 is a flow chart for explaining the correlation method;

[0076] FIG. 12 is a schematic diagram showing the construction of the first embodiment of a mark position detecting system according to the present invention;

[0077] FIG. 13 is an overview flow chart for explaining the mark position detecting method in the first embodiment of a method for detecting a mark position according to the invention;

[0078] FIGS. 14 and 15 are flow charts for explaining the mark position detecting method in the first embodiment of a method for detecting a mark position according to the invention;

[0079] FIGS. 16A and 16B are diagrams showing examples of measured waveforms obtained at steps shown in FIGS. 14 and 15 ;

[0080] FIG. 17 is a schematic diagram showing the construction of the second embodiment of the mark position detecting system according to the present invention;

[0081] FIG. 18 is an overview flow chart for explaining the mark position detecting method in the second embodiment of a method for detecting a mark position according to the invention;

[0082] FIGS. 19 through 21 are flow charts for explaining the mark position detecting method in the second embodiment of a method for detecting a mark position according to the invention; and

[0083] FIGS. 22A and 22B are diagrams showing waveforms of diffracted rays of reflected beam denoting wavelength dependency of the reflected beam from a measurement mark.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0084] Referring now to the accompanying drawings, some preferred embodiments of the present invention will be described below.

[0085] (1) First Preferred Embodiment

[0086] FIG. 12 is a schematic diagram showing the construction of the first embodiment of a mark position detecting system according to the present invention.

[0087] As shown in FIG. 12 , the mark position detecting system 1 of the present embodiment comprises an illumination optical system 10 , a stage 70 , a detection optical system 30 and a control computer 50 .

[0088] The stage 70 supports thereon a substrate 120 which is provided with a misalignment measurement mark as an object to be measured.

[0089] The illumination optical system 10 includes a light source 13 and a half mirror 15 . The light source 13 emits light L 1 having an arbitrary wavelength λ on the basis of a command supplied from the control computer 50 .

[0090] The detection optical system 30 includes a CCD sensor 33 and an A/D (analogue to digital) converter 35 . The CCD sensor 33 detects the reflected beam L 2 from the misalignment measurement mark 20 , and generates a signal of electric charges according to the strength of the detected beam. The A/D converter 35 receives the analogue signal supplied from the CCD sensor 33 to convert the signal to a digital signal.

[0091] The control computer 50 includes a measured waveform diagram preparing part 51 , a theoretical waveform preparing part 57 , a waveform comparison part 53 and a misalignment operation part 55 .

[0092] The waveform diagram preparation part 51 receives the digital signal from the A/D converter 35 to prepare a waveform diagram of the measured beam on the basis of the received digital signal. The waveform diagram has a horizontal axis and a vertical axis, the horizontal axis denoting position coordinates of the substrate 120 in the X direction and the vertical axis denoting strengths of the reflected beam.

[0093] The theoretical waveform preparing part 57 receives information Inf concerning a material and a shape of the misalignment measurement mark, calculates strength of a reflected beam corresponding to that from a desired position of the measurement mark by means of simulation and prepares a theoretical waveform. In this embodiment, information Inf on the misalignment measurement mark includes the material and the thickness T 1 of the SiO 2 film (the first film), each height of steps (depths) D 1 , D 2 of the first and second concavities C 1 , C 2 , and the material and the thickness T 2 of the SiN film (the second film).

[0094] The waveform comparison part 53 compares the measured waveform with the theoretical waveform, detects a portion of the measured waveform which is most similar to the theoretical waveform and output a position coordinate of the measured waveform corresponding thereto.

[0095] The misalignment operation part 55 carries an arithmetical operation on the positional information supplied from the waveform comparison part 53 , determines whether there is an error in the measured result or not and outputs the value when any error exists.

[0096] Referring to the accompanying drawings, as a preferred embodiment of a mark position detecting method according to the present invention, the operation of the mark position detecting system 1 of the above-mentioned embodiment will be described below. In addition, the same mark as the measurement mark shown in FIGS. 2 and 7 will be used as a misalignment measurement mark in the following description for easy comparison with the conventional methods.

[0097] FIGS. 13 through 15 are flow charts for explaining the mark position detecting method in the present embodiment. FIG. 13 is an overview flow chart showing a series of steps in the embodiment, and FIGS. 14 and 15 shows a series of main steps which will be described later.

[0098] In this preferred embodiment the thickness of films and the depths of the concavities of the measurement mark are previously measured as a pre-process prior to detection of the position of the measurement mark. Steps S 11 and S 12 in FIG. 13 denote the pre-process prior to the measurement. Meanwhile, the step S 40 ( FIGS. 14 and 15 ) denotes the maim process for the measurement. The pre-process and the main process are separately described below.

[0099] (A) Pre-Process

[0100] First, as shown in FIG. 13 , with respect to the measurement mark shown in FIG. 2 or 7 , the film thickness T 1 (the first film thickness) of the SiO 2 layer 23 and the film thickness T 2 (the second film thickness) of the SiN layer 27 is measured with a thicknessmeter (not shown)(step S 11 ).

[0101] Then, depths of first and second concavities of the measurement mark are measured with a step measuring instrument (a thicknessmeter, not shown) (step S 12 ). With respect to the symmetric measurement mark 20 shown in FIG. 2 , the first and second depths correspond to the depths D 1 and D 2 of the concavities C 1 and C 2 respectively. With respect to the measurement mark 21 shown in FIG. 7 , the first and second depths correspond to the depths D 1 ′ and D 2 ′ of the concavities C 3 and C 4 respectively.

[0102] (B) Main Process

[0103] A position of a misalignment measurement mark is then detected using data on the film thickness and the depths obtained the above-mentioned pre-process (step S 40 ).

[0104] First, as shown in FIG. 14 , light L 1 having a wavelength λ is emitted from the light source 13 of the mark position measurement system 1 to illuminate the misalignment measurement mark via the half mirror 15 (step S 41 ). light L 1 is then deflected on the misalignment measurement mark to be the reflected beam of light L 2 . The reflected beam L 2 passes through the half mirror 15 again to be incident on a pixel part of the CCD sensor 33 as a beam of light having strength dependent on the difference between optical path lengths of rays consisting the beam.

[0105] The CCD sensor 33 detects the reflected beam L 2 . That is, electric charges are generated in accordance with the strength of the reflected beam incident on each pixel part (step S 42 ) and the CCD sensor 33 supplies signals of the charges to the A/D converter 35 as analogue signals.

[0106] The A/D converter 35 converts the signals supplied from the CCD sensor 33 to digital signals and supplies the converted signals to the control computer 50 . The measured waveform preparing part 51 in the control computer 50 recognizes strength distribution of the reflected beam dependent on a change of the surface of the misalignment measurement mark on the basis of the supplied signals to prepare a measured waveform diagram showing a measured waveform WSM (step S 43 ). The measured waveform diagram corresponds to a diagram which has a horizontal axis and a vertical axis and is plotted with measured values. The horizontal axis denotes position coordinates in X directions of a plane parallel to the surface of the substrate 120 having the measurement mark and the vertical axis denotes the strengths of the reflected beam from the measurement mark.

[0107] Steps of detecting a position of a misalignment measurement mark by processing the measured waveform WSM will be described below separately in a case in which a measurement mark has a symmetrical cross sectional shape such as the measurement mark 20 shown in FIG. 2 and a case in which a measurement mark has a non-symmetrical cross sectional shape such as the measurement mark 21 shown in FIG. 7 .

[0108] (a) When Cross Sectional Shape Of Measurement Mark Is Symmetric

[0109] An example of the measured waveform WSM 1 obtained at the above mentioned steps S 41 through S 43 is shown in FIG. 16A . Like the waveform shown in FIG. 4 , the measured waveform in FIG. 16A has a symmetrical shape with respect to a line passing through the center of the SiN layer 27 and perpendicular to the substrate 120 in cross sectional view of the measurement mark 20 . A method for detecting a mark position using the symmetrical waveform WSM 1 is described below referring to flow charts of FIGS. 14 and 15 , and FIG. 16A . For a simplified explanation, in FIGS. 14 and 15 , a film thickness is designated by Tj (J=1, 2), a depth is designated by Dk (k=1, 2) and a position coordinate of the measurement mark by X 1 , Xm (l, m=1 to 6).

[0110] First, as shown in FIG. 14, a theoretical waveform is prepared by simulation to detect a position on the measured waveform diagram corresponding to a position coordinate X 1 .

[0111] That is, j=1, k=1, l=1 and m=2 are set (step S 44 ), a film thickness T 1 and a depth D 1 of a step are inputted into the theoretical waveform preparation part 57 in the control computer 50 as parameters. In this preferred embodiment T 1 =1 μm and D 1 =0.1 μm are inputted. The theoretical waveform preparation part 57 calculates light strength of the reflected beam which is expected to be generated at and near the position coordinate X 1 on the basis of the inputted parameters to prepare a theoretical waveform WST 21 as shown in FIG. 16 A and supplies the theoretical waveform WST 21 to the waveform comparison part 53 (step 45 in FIG. 14 ).

[0112] The waveform comparison part 53 then compares the measured waveform WSM 1 with the theoretical waveform WST 21 , detects a portion of the measured waveform WSM 1 which has position coordinates from X 1 to X 2 and near thereto and is most similar to the theoretical waveform WST 21 . The waveform comparison part 53 further detects a position coordinate XP 1 corresponding to the point at which the strength of the reflected light drops (step S 46 in FIG. 14 ) and supplies the information on the XP 1 to the misalignment operation part 55 .

[0113] Next, at the same steps as the above-mentioned steps the waveform comparison part 53 detects a position XP 6 on the measured waveform corresponding to a position coordinate X 6 (steps S 47 , S 48 , S 45 and S 46 ). That is, j=1, k=2, l=6 and m=5 are set (step S 48 ), T 1 =1 μm and D 2 (=D 1 )=0.12 μm are inputted to the theoretical waveform preparation part 57 as parameters. The theoretical waveform preparation part 57 prepares a theoretical waveform WST 26 of X 6 and position coordinates in the vicinity of X 6 (step S 45 ). The waveform comparison part 53 then detects the position coordinate XP 6 of the portion of the measured waveform WSM 1 which has position coordinates from X 6 to X 5 and near thereto and is most similar to the theoretical waveform WST 26 (step S 46 ). The waveform comparison part 53 then supplies the information on the XP 6 to the misalignment operation part 55 .

[0114] After XP 1 and XP 6 are detected (step S 47 ), as shown in FIG. 15 , the misalignment operation part 55 calculates the middle point XP 16 of XP 1 and XP 6 (step S 49 ).

[0115] Then, XP 3 and XP 4 corresponding to position coordinates X 3 and X 4 respectively are detected, meanwhile X 3 and X 4 are the position of the outside edges of the SiN layer 27 respectively. These steps are substantially the same as the above-mentioned steps S 45 and S 46 .

[0116] In more specific, as shown in FIG. 15 , j=1, k=2, l=3 and m=4 are set (step S 51 ) and the film thickness Ti of the SiO 2 layer 23 and the film thickness of SiN film T 2 are inputted to the theoretical waveform preparation part 57 as parameters. The theoretical waveform preparation part 57 calculates light strength of a reflected beam which is expected to be obtained at and near the position coordinate X 3 to prepare a theoretical waveform WST 23 (step S 52 ). In this preferred embodiment T 2 is 1 μm.

[0117] The waveform comparison part 53 then detects the position coordinate XP 3 of the portion of the measured waveform WSM 1 which has the position coordinates from X 3 to X 4 and near thereto and is most similar to the theoretical waveform WST 23 (step S 53 ).

[0118] At similar steps, with respect to XP 4 (step S 54 ), j=1, k=2, l=4 and m=3 are set (step S 55 ), XP 4 is detected at above-mentioned steps and the detected result is supplied to the misalignment operation part 55 .

[0119] The misalignment operation part 55 first calculates the middle point XP 34 of XP 3 and XP 4 (step S 56 ) and further calculates the difference between the middle point XP 16 obtained at the step S 49 and the middle point XP 34 , i.e. (XP 16 −XP 34 ) to output the calculated value as misalignment “0”.

[0120] As described above, when the misalignment measurement mark having a symmetric cross sectional shape is to be detected in this preferred embodiment, a theoretical waveform corresponding to a surface region which changes in shape, such as WST 21 and WST 26 is prepared and the position coordinates XP 1 and XP 2 of a portion of the measured waveform, which portion is most similar to the theoretical waveform respectively, is acquired. Each of the position coordinates XP 1 and XP 6 thus accurately coincides with the wafer position coordinates X 1 and X 6 of the measurement mark respectively as long as the theoretical waveforms WST 21 and WST 26 are exactly prepared. Since XP 16 precisely coincides with the middle point of that of the wafer position coordinates X 1 and X 6 , the misalignment measurement mark can be accurately detected in the preferred embodiment.

[0121] (b) When Cross Sectional Shape Of Measurement Mark Is Non-Symmetric

[0122] Next, steps of detecting a measurement mark when a cross sectional shape thereof is non-symmetric will be described below referring to FIGS. 14, 15 and 16 B. The mark 21 shown in FIG. 7 will be used as an object to be measured. Then, the film thickness T 1 =T 2 =1 μm, the depth D 1 ′=0.1 μm of the first concavity and the depth D 2 ′=0.14 μm of the second concavity are measured at step S 41 shown in FIG. 14 and values thereof are inputted to the theoretical waveform preparing part 57 .

[0123] FIG. 16B shows an example of a measured waveform WSM 2 obtained at the above mentioned steps S 41 through S 43 . As mentioned earlier, the measured waveform WSM 2 has a non-symmetrical shape with respect to a line passing through the center of the SiN layer 27 and perpendicular to the substrate 120 in cross sectional view of the measurement mark 21 .

[0124] First, as shown in FIG. 14, a theoretical waveform which is expected to be obtained from position coordinates of and near X 1 is prepared to detect a position on the measured waveform diagram corresponding to the position coordinate X 1 . That is, j=1, k=1, l=1 and m=2 are set (step S 44 ), the film thickness T 1 =1 μm and the depth D 1 ′=0.1 μm of the step are inputted to the theoretical waveform preparation part 57 in the control computer 50 as parameters. The theoretical waveform preparation part 57 calculates light strength of a reflected beam which is expected to be generated at and near the position coordinate X 1 to prepare a theoretical waveform WST 21 ′ as shown in FIG. 16B (step 45 ).

[0125] The waveform comparison part 53 then compares the measured waveform WSM 1 with the theoretical waveform WST 21 ′, detects a portion of the measured waveform WSM 1 which has the position coordinates from X 1 to X 2 and near thereto and is most similar to the theoretical waveform WST 21 ′. The waveform comparison part 53 further detects the position coordinate XP 1 corresponding to the point at which strength of the reflected light drops (step S 46 in FIG. 14 ) and supplies the information on the XP 1 to the misalignment operation part 55 .

[0126] Next, at the same steps as the above-mentioned steps a theoretical waveform which would be obtained on a reflected beam from the region having position coordinates of and near X 6 and the position XP 6 on the measured waveform corresponding to X 6 is detected (steps S 47 , S 48 , S 45 and S 46 ). That is, j=1, k=2, l=6 and m=5 are set (step S 48 ), T 1 =1 μm and D 2 ′ (≠D 1 ′)=0.14 μm are inputted to the theoretical waveform preparation part 57 as parameters. The theoretical waveform preparation part 57 prepares a theoretical waveform WST 26 ′ of and near X 6 (step S 45 ). The waveform comparison part 53 then detects the position coordinate XP 6 of the portion of the measured waveform WSM 2 which has the position coordinates from X 6 to X 5 and near thereto and is most similar to the theoretical waveform WST 26 ′ (step S 46 ). The waveform comparison part 53 then supplies the detected result to the misalignment operation part 55 .

[0127] After XP 1 and XP 6 are detected (step S 47 ), as shown in FIG. 15 , the misalignment operation part 55 calculates the middle point XP 16 of XP 1 and XP 6 (step S 49 ).

[0128] Then, XP 3 and XP 4 corresponding to the position coordinates X 3 and X 4 respectively are detected, meanwhile X 3 and X 4 are the position of the outside edges of the SiN layer 27 respectively.

[0129] In more specific, as shown in FIG. 15 , j=1, k=2, l=3 and m=4 are set (step S 51 ) and the film thickness T 1 =1 μm of the SiO 2 layer 23 and the film thickness of SiN layer T 2 =1 μm are inputted to the theoretical waveform preparation part 57 as parameters. The theoretical waveform preparation part 57 calculates light strength of a reflected beam which is expected to be obtained at and near the position coordinate X 3 to prepare the theoretical waveform WST 23 ′ (step S 52 ).

[0130] The waveform comparison part 53 then detects the position coordinate XP 3 of the portion of the measured waveform WSM 1 which has the position coordinates from X 3 to X 4 and near thereto and is most similar to the theoretical waveform WST 23 ′ (step S 53 ).

[0131] At similar steps, with respect to XP 4 (step S 54 ), j=1, k=2, l=4 and m=3 are set (step S 55 ), XP 4 is detected at above-mentioned steps S 52 and S 53 , and the detected result is supplied to the misalignment operation part 55 .

[0132] The misalignment operation part 55 first calculates the middle point XP 34 of XP 3 and XP 4 (step S 56 ) and further calculates the difference between the middle point XP 16 obtained at the step S 49 and the middle point XP 34 , i.e. (XP 16 −XP 34 ) to output the calculated value as misalignment “0”.

[0133] In the present embodiment as described above, a theoretical waveform corresponding to a surface region which changes in shape is prepared and the position coordinates in a portion of the measured waveform WSM, such as XP 1 and XP 2 , which portion is most similar to the theoretical waveform, is acquired. For this reason the position coordinates XP 1 and XP 6 accurately coincide with the wafer position coordinate X 1 and X 6 of the measurement mark respectively as long as the theoretical waveform is exactly prepared.

[0134] As mentioned above in specific, according to the present embodiment a misalignment measurement mark can be accurately detected whether a shape of measured waveform mark is symmetric or non-symmetric.

[0135] (2) Second Preferred Embodiment

[0136] Referring to the accompanied drawings the second preferred embodiment of the present invention will be described below.

[0137] FIG. 17 is a schematic diagram showing the construction of the second embodiment of a mark position detecting system according to the present invention. As shown in this figure, the mark position detecting system 2 of the present embodiment is characterized in that the system further comprises a shape information acquiring part for measuring the film thickness T 1 , T 2 and the depths D 1 , D 2 of the concavities which are parameters used for detecting the measurement mark.

[0138] As can be seen in comparison with FIG. 12 , the illumination optical system 11 of the mark position detecting system 2 further includes a half mirror 17 . In addition, a detection optical system 31 further includes a spectroscope 41 and an A/D converter 47 . Furthermore, a control computer 60 further includes parameter calculation part 63 . These additional component parts constitute the shape information acquiring part. Other components of the mark position detecting system 2 of this preferred embodiment are substantially the same as the mark measurement detecting system 1 shown in FIG. 12 .

[0139] The spectroscope 41 has a diffraction grating 43 and a CCD sensor 45 . A reflected beam of light from the measurement mark travels through the half mirror 17 to be incident on the diffraction grating 43 . The incident beam of light is broken up into rays having arbitrary wavelengths by the diffraction grating 43 to be incident on a pixel part of the CCD sensor 45 .

[0140] Charges generated in each pixel of the pixel part of the CCD sensor 45 are outputted as an analogue signal. The analogue signal is supplied to the A/D converter 47 to be converted to a digital signal. The converted digital signal is supplied to the control computer 60 as a signal decoding to strength of each ray having each wavelength separated from the reflected beam.

[0141] Referring now to FIGS. 18 through 22 A and 22 B, as a preferred second embodiment of a mark position detecting method according to the present invention, the operation of the mark position detecting system 2 of this embodiment will be described below.

[0142] FIGS. 18 through 21 are flow charts for explaining the mark position detecting method of the preferred second embodiment. FIG. 18 is an overview flow chart for explaining the mark position detecting method and FIGS. 19 through 21 are flow charts for explaining a step S 20 of the FIG. 18 in more specific.

[0143] As shown in FIG. 18 , the present embodiment is characterized in that the method includes a step of calculating a parameter as a preparation step (step S 20 ) prior to a misalignment measurement mark position detecting step (step S 40 ) as the main step. Parameters used in this embodiment are the film thickness T 1 , T 2 of the measurement mark and the depths D 1 , D 2 of the concavities shown in FIG. 2 or FIG. 7 . Information on materials constructing each thin film is previously stored in a memory (not shown) of the control computer 60 . Since the misalignment measurement mark position detecting step (step S 40 ) as a main step is the same as the steps shown in FIGS. 13 through 15 , the preparation step will be described in detail below with an example of the non-symmetric measurement mark 21 . The flow chart of FIG. 19 shows steps of calculating the film thickness T 1 and the depth D 1 , and the flow chart of FIG. 20 shows steps of calculating the depth D 2 . In addition, the flow chart of FIG. 21 shows steps of calculating the film thickness T 2 .

[0144] First, as shown in FIG. 19 , the parameter calculation part 63 in the control computer 60 sets values T 1 a, T 1 b and T 1 c, and D 1 a, D 1 b and D 1 as candidate values for film thickness T 1 and depth D 1 respectively, and supplies these candidate values to the theoretical waveform preparation part 57 (step S 21 ).

[0145] The theoretical waveform preparation part 57 extracts information on the materials of the SiO 2 layer 23 and SiN layer 27 from the memory (not shown). The theoretical waveform preparation part 57 calculates wavelength dependency of strength of the reflected beam by means of simulation, which would be obtained if the first concavity (position coordinates X 1 to X 2 ) of the measurement mark is illuminated with the white light, on the basis of the extracted information and the supplied parameters. The theoretical waveform preparation part 57 then prepares theoretical waveforms WST 31 a, WST 31 b and WST 31 c which correspond to each of the candidate values T 1 a and D 1 a, T 1 b and D 1 b, and T 1 c and D 1 c as shown in FIG. 22A (step S 22 ).

[0146] Next, light having wavelength of λ, in this preferred embodiment white light L 1 is emitted from the light source 13 to irradiate the first concavity (position coordinates X 1 to X 2 ) of the measurement mark via the half mirror 15 such as pointed by an arrowhead L 1 a shown in FIG. 2 (step S 23 ). The reflected beam of light enters the spectroscope 41 via the half mirror 17 and diffracted by the diffraction grating 43 into rays having predetermined wavelengths respectively. The diffracted rays are incident on the pixel part of the CCD sensor 45 and electric charges in accordance with strength of each of the diffracted rays are generated every diffracted ray in each pixel and a signal of the charges are outputted (step S 24 ). In this embodiment wavelengths of the diffracted rays are set as 600 nm, 650 nm, 700 nm, 750 nm and 800 nm and the strength of these rays are measured. Signals of the charges generated in the CCD sensor 45 are converted to digital signals by the A/D converter 47 to be supplied to the theoretical waveform preparing part 57 . The theoretical waveform preparing part 57 prepares a measured waveform of the diffracted rays WSM 31 denoting wavelength dependency of the reflected rays on the basis of the signals supplied from the A/D converter 47 . In the example shown in FIG. 22 A, for simple explanation, the measured value of the strength of each ray wavelength is calculated to be plotted in a theoretical waveform diagram which is previously prepared at the above mentioned step 22 . The theoretical waveform preparing part 57 supplies information on the prepared waveform of the measured diffracted rays to the waveform comparison part 53 (step S 25 ).

[0147] The waveform comparison part 53 compares the waveform of the measured diffracted rays (measured values of each wavelength of the rays of reflected beam in the example shown in FIG. 22A ) with the theoretical waveform, select the waveform which is most similar to the waveform of the measured diffracted rays (measured values), of the theoretical waveforms WST 31 a, WST 31 b and WST 31 c. The waveform comparison part 53 supplies information on the selected waveform to the parameter calculation part 63 . In the example shown in FIG. 22A the theoretical waveform WST 31 b is selected.

[0148] A specific method for comparing the measured waveform of the diffracted rays with the theoretical waveform is described below.

[0149] That is, the ray strength of the theoretical waveform in each wavelength are assumed as a (λ), b (λ) and c (λ), and the ray strength of the measured diffracted rays in each wavelength are assumed as Y (λ).

[0150] Then, the total sum of the absolute values of the difference between the ray strength in the theoretical waveforms and the strength of the ray of the reflected beam which are measured, that is,

Σ|Y(λ)−a(λ)|, Σ|Y(λ)−b(λ)|, Σ|Y(λ)−c(λ)| (3)

[0151] is calculated and the theoretical waveform which gives the minimum value according to the formula (3) is selected as a theoretical waveform which is most similar to the measured diffracted waveform (measured value).

[0152] The parameter calculation part 63 selects the film thickness T 1 and the depth D 1 which correspond to the values of the theoretical waveform WST 31 b on the basis of the information on the theoretical waveform WST 31 b supplied from the waveform comparing part 53 as parameters to supply the values to the theoretical waveform preparing part 57 (step S 26 ). In the example shown in FIG. 22A T 1 =1 μm and D 1 =0.1 μm are selected.

[0153] Then, the mark position detecting system 2 measures the depth D 2 of the concavity C 2 of the measurement mark.

[0154] That is, as shown in FIG. 20 , the parameter calculation part 63 sets values D 2 a, D 2 b and D 2 c as candidate values for depth D 2 , and supplies these candidate values to the theoretical waveform preparation part 57 (step S 27 ).

[0155] The theoretical waveform preparation part 57 extracts information on the materials of the SiO 2 layer 23 from the memory (not shown). The theoretical waveform prepara