A device and method for detecting distortion and magnification of a multi-channel objective lens

Abstract

The present invention discloses a device and method for detecting the distortion and magnification of a multi-channel objective lens. The device comprises an illumination system, a mask stage, a test reticle, a projection objective lens, a workpiece stage, and a multi-channel image plane sensor. The multi-channel image plane sensor simultaneously detects the image position offset between the actual image points and the nominal image points of multiple object surface test marks after imaging through the projection objective lens, and calculates the distortion and magnification errors of the objective lens through fitting. This shortens the detection time, eliminates the influence of the workpiece stage error on the detection accuracy, and improves the detection accuracy.

Description

Technical Field

[0001] The present invention belongs to the technical field of photolithography, and in particular provides a device and method for detecting distortion and magnification error of a multi-channel photolithography projection objective lens. Background Art

[0002] The distortion of lithography projection lenses is one of the most important factors affecting the overlay accuracy of lithography machines. Distortion causes the lateral magnification of the lens to change as the field of view increases, causing the pattern exposed on the silicon wafer to shift relative to its ideal position, resulting in overlay errors. To control the imaging quality (image quality) of the projection lens, distortion needs to be detected and optimized during the installation and use of the lens, requiring high-precision image quality detection technology. The imaging quality is tested offline during the integrated assembly stage of the projection lens; the imaging quality is tested in situ during the use of the lens. Based on the test results, aberrations can be compensated to improve the imaging quality by adjusting the movable lens of the projection lens.

[0003] The principle for detecting distortion in lithography projection lenses is to accurately measure the deviation between the actual and nominal image positions of a set of test marks on the object surface after imaging through the projection objective. The least-squares method is then used to fit the distortion and magnification error of the projection objective. Distortion detection techniques for lithography projection lenses can be categorized into three types: exposure detection, aerial image detection, and wavefront detection, depending on the method used to obtain the offset of the image positions of the test marks on the object surface.

[0004] Prior art CN102466977B proposes an exposure-based distortion detection device, comprising an illumination system, a mask stage, a reticle, a projection objective, a silicon wafer stage, and a silicon wafer. The distortion detection method involves exposing two layers of mask patterns on a silicon wafer, measuring the overlay error, and fitting the distortion. When exposing the first layer, the array of measurement marks on the mask is exposed onto the silicon wafer through the entire field of view of the projection objective. The actual imaging position of the marks is affected by distortion. When exposing the second layer, the illumination area is limited to a small area on the mask, so that only the marks in the center field of view are illuminated. The wafer stage is then moved to the nominal imaging position of the first layer's exposed pattern, and the second layer's pattern is exposed, with the small squares of the second layer's pattern aligning with the nominal imaging position of the large squares of the first layer's exposure, forming overlay marks. Since each mark on the second layer is derived from the marks exposed in the center field of view, the actual imaging position is unaffected by distortion. Therefore, the offset in the imaging position of the overlay marks, i.e., the overlay error, is the distortion. A disadvantage of this method is that the fitting results are affected by stage errors. During the second layer exposure, the workpiece stage positioning error varies at each field of view. To achieve the desired fit, it is usually assumed that the workpiece stage error is the same across all fields of view during the second layer exposure. Therefore, different workpiece stage errors will be reflected in the distortion fitting values, resulting in inaccurate fitting results. Furthermore, the accuracy of distortion detection is also limited by the accuracy of overlay error measurement.

[0005] US20020041377A1 proposes a distortion detection technology based on aerial images. This method involves placing an aerial image sensor on a silicon wafer surface, sequentially scanning the aerial image of each measurement mark, determining the actual imaging position of all aerial images, and fitting the distortion. This technology requires moving the aerial image sensor to measure the position of each measurement mark's aerial image. The positioning error of the workpiece stage varies between measurements, making it impossible to fit the distortion. To ensure this, the positioning error of the workpiece stage is typically assumed to be the same for each measurement. As a result, the distortion obtained through fitting is inaccurate, and the accuracy of this technology is affected by the workpiece stage's positioning error.

[0006] Document CN102540751A proposes a distortion detection technology based on wavefront detection. The method involves installing a mask containing a pinhole array on a mask stage and a Hartmann sensor on a workpiece stage. Light passes through the pinholes to produce an ideal wavefront. The ideal wavefront, after passing through the projection objective, carries aberrations in the wavefront to be measured. The wavefront to be measured is collimated by a collimator and received by a Hartmann sensor. The Hartmann sensor is aligned with the spatial image of each pinhole mark to measure its wavefront error. Based on the Z2 to Z4 terms of the wavefront error measured by the Hartmann sensor, the actual imaging point position of the pinhole mark can be calculated and fitted to obtain the distortion. The distortion detection accuracy of this technology is also affected by the positioning error of the workpiece stage.

[0007] The accuracy of the above three distortion detection technologies are all affected by the positioning error of the workpiece stage. Some of them have slow measurement speed, complex test steps, and cannot measure the wavefront aberration and distortion of the projection objective lens at the same time. In addition, these technologies cannot measure the positioning error of the workpiece stage. Summary of the Invention

[0008] The object of the present invention is to provide a device and method for detecting distortion and magnification of a multi-channel objective lens. The technical solution of the present invention is as follows:

[0009] In a first aspect, the present invention provides a multi-channel objective lens distortion and magnification detection device, comprising:

[0010] A lighting system for providing light sources;

[0011] Mask stage, used to carry test masks;

[0012] A test mask, comprising A (A>=1) groups, each group having G (G>=3) object plane test marks, wherein the G object plane test marks in each group have the same pitch p and spatial distribution, and are arranged along a straight line, a broken line, or an array according to the pitch p, and the pitch between different groups is greater than or equal to p;

[0013] A projection objective lens, used for imaging the test marks on the object surface;

[0014] A multi-channel image plane sensor is used to simultaneously detect the imaging position offset between the actual image points and the nominal image points of multiple object plane test marks after imaging through a projection objective lens; it comprises a group of G image plane test marks and an array detector that are conjugate with a group of G object plane test marks, and the image plane test marks and the array detector are arranged in sequence along the light transmission direction; the spacing between the G image plane test marks is p'=M×p, where M is the nominal micro-magnification of the projection objective lens; the divergent light transmission distance between the array detector and the plane where the image plane test marks are located is q, and the divergent light transmission distance is q. This ensures that the optical signals of different marks on the array detector of the multi-channel image plane sensor do not interfere with each other;

[0015] The workpiece stage is used to carry the multi-channel image sensor.

[0016] The multi-channel image plane sensor is a multi-channel point diffraction interferometer sensor, a multi-channel Shack-Hartmann sensor or a multi-channel shearing interferometer sensor.

[0017] The multi-channel point diffraction interferometer sensor includes an image plane test mark, a collimating lens array and an area array detector. The object plane test mark includes two pinhole marks, the image plane test mark includes a pinhole mark and a light window, the collimating lens array includes G collimating lenses, and the front focal plane of the collimating lens array is located in the plane where the image plane test mark is located. The area array detector is a CCD, and the photosensitive surface of the area array detector is located at the back focal plane of the collimating lens array. The focal lengths of the G collimating lenses are equal, and the focal length is f1. This ensures that light signals of different marks on the area array detector of the multi-channel image sensor do not interfere with each other.

[0018] The multi-channel Shack-Hartmann sensor includes an image plane test mark, a collimating lens array, a microlens array, and an area array detector. The object plane test mark is a pinhole mark, the image plane test mark is a light-transmitting window, the collimating lens array includes G collimating lenses, the front focal plane of the collimating lens array is located in a plane conjugate with the object plane test mark, the area array detector is a CMOS, the microlens array is located in the back focal plane of the collimating lens array, the photosensitive surface of the area array detector is located in the focal plane of the microlens array, the focal length of the G collimating lenses is equal, and the focal length is f2. This ensures that light signals of different labels on the area array detector of the multi-channel Shack-Hartmann sensor do not interfere with each other.

[0019] The multi-channel shearing interferometer sensor includes an image plane test mark and an array detector. The object plane test mark includes one-dimensional gratings in two orthogonal directions, which are called the first direction and the second direction. The one-dimensional gratings in the first direction and the second direction are used to measure the imaging position offset in the x direction and the y direction, respectively. The grating size and grating period in each direction are the same. The image plane test mark is a checkerboard grating, which is located on a plane conjugate to the object plane test mark. The array detector is a CCD, which is located on the side of the checkerboard grating away from the projection objective lens. The distance between the plane where the checkerboard grating is located and the photosensitive surface of the array detector is This ensures that light signals of different marks on the area array detector of the multi-channel shearing interferometer sensor do not interfere with each other.

[0020] In a second aspect, the present invention further provides a method for testing the distortion and magnification of a multi-channel objective lens, which is performed using the multi-channel objective lens distortion and magnification detection device, and is characterized in that it includes the following steps:

[0021] Step 1: Install the test mask on the mask stage and move the object plane test mark into the field of view of the projection objective lens;

[0022] Step 2: Install the multi-channel image plane sensor on the workpiece table and move the multi-channel image plane sensor to a plane that is conjugate with the object surface test mark;

[0023] Step 3: Use the lighting system to image the test marks on the object surface through the projection objective lens, and use the multi-channel image plane sensor to simultaneously image the first group of G test marks on the object surface through the projection objective lens. i ,Δy i )(I=1, i=1~G) is detected, and the imaging position offset of G channels can be detected each time. The field of view is moved to measure the imaging position offset of the next group of object surface test marks in the field of view. There are two ways to move the field of view: a. Change the setting of the lighting system so that the light source it emits only illuminates the second group of G object surface test marks, move the workpiece stage, and use a multi-channel image plane sensor to simultaneously measure the imaging position offset (Δx i ,Δy i )(I=2,i=1~G) to detect, and so on, until the imaging position offset (Δx i ,Δy i b. Move the mask stage to change the position of the field point where the object surface test mark is located, and move the workpiece stage so that the multi-channel image plane sensor can once again simultaneously image the first group of G object surface test marks through the projection objective lens, and the imaging position offset (Δx i ,Δy i)(I=2,i=1~G) to detect, and so on, until the imaging position offset (Δx i ,Δy i );

[0024] Step 4: (Δx i ,Δy i ) According to formula (1), the least squares fitting method is used to calculate the distortion and magnification error of the projection objective lens:

[0025]

[0026] , where T Ix 、T Iy are respectively the translation errors of the workpiece stage (900) when the multi-channel image plane sensor (600) is used to detect the imaging position offsets of the object plane test marks (300) in the x-direction and the y-direction of the Ith (I=1, 2...A)th group, θ Ix ,θ Iy are respectively the rotation errors of the workpiece stage (900) when the multi-channel image plane sensor (600) is used to detect the imaging position offsets of the object plane test marks (300) in the x-direction and the y-direction of the I (I=1, 2...A) group. x 、M y is the magnification of the projection objective lens (500) in the x-direction and the y-direction, is the distance from the nominal imaging position of the i-th object plane test mark (300) to the center of the image plane coordinate system, k1 is the third-order distortion coefficient of the projection objective lens (500), r x 、r y is the fitting residual, (x i ,y i ) is the nominal imaging position of the i-th (i=1, 2...G) object plane test mark (300) in the I-th group, and the formula is as follows:

[0027] (x i ,y i )=(xx i ,yy i )·M (2)

[0028] , where (xx i ,yy i ) is the position of the object surface test mark (300) in the object surface coordinate system. For N object surface test marks, there are 2N equations in total, and the required unknowns are the workpiece stage errors (T Ix ,θ Ix 、T Iy ,θ Iy )(I=1~A) and a set of third-order distortion and magnification error of projection objective lens (k1, Mx 、M y ), assuming that the number of unknowns in the equation group (1) is H, 3A+3<=H<=4A+3, when the number of equations in the equation group (1) is greater than or equal to the lower limit of H, that is, when equation (6) is satisfied, the fitting of the third-order distortion coefficient and magnification error parameters of the measured projection objective can be achieved:

[0029]

[0030] , since A>=1, therefore G>=3.

[0031] In step 3, different multi-channel image sensors are used to test the imaging position offset (Δx i ,Δy i )(I=1~A, i=1~G) steps are different.

[0032] The test steps when using a multi-channel point diffraction interferometer are as follows: the illumination light source passes through the first set of G double pinholes on the object plane test mask to generate 2G ideal spherical waves, which are then imaged by the projection objective lens to form 2G measured wavefronts. The workpiece stage is moved, and when the G image pinhole marks of the multi-channel point diffraction interferometer are aligned with the nominal imaging positions of the G object plane pinholes, the position of the wafer stage at this time is recorded [X I , Y I , Z I ](I=1), one path (G) of the 2G measured wavefronts is diffracted by the G image plane pinhole marks to form an ideal spherical wave as the reference light, and one path (G) passes through the optical window as the measurement light. The G measurement lights and the G reference lights interfere with each other and are collimated by the G collimating lenses, resulting in G interference fringes on the CCD. The interference fringes are analyzed to obtain the wavefront errors of the first group of G field points. Based on the Zernike coefficients of the measured G wavefront errors, the Z2 to Z4 terms are substituted into formula (3) to calculate the alignment position deviation (δx) of the first group of G pinholes. i ,δy i ,δz i ):

[0033]

[0034] Among them, Rsensor is the wavefront radius detected by CCD, f is the focal length of the collimating lens in the multi-channel point diffraction interferometer sensor, (δx i ,δy i ,δz i ) is the position deviation between the actual image point of each pinhole mark in the first group and the reference point (the focus of the collimating lens of the point diffraction interference sensor), and (δx i ,δy i ,δz i), add the wafer stage position [X I , Y I , Z I ](I=1), which is the actual imaging point position (x ri ,y ri , z ri ), the nominal imaging position (x i ,y i ), then the imaging position offset of the first group G object plane pinhole marks (Δx i ,Δy i The calculation formula of )(I=1, i=1~G) is formula (4):

[0035] (Δx i ,Δy i )=(x ri ,y ri )-(x i ,y i ) (4)

[0036] Move the field of view and measure the imaging position offset (Δx i ,Δy i )(I=2, i=1~G), until the imaging position offset (Δx i ,Δy i )(I=1~A, i=1~G).

[0037] The test steps of the multi-channel Shack-Hartmann sensor are as follows: the light emitted by the illumination system passes through the first group of G object pinholes to form G ideal spherical waves, which are imaged by the projection objective lens to form G measured wavefronts. The multi-channel Shack-Hartmann sensor is moved until the front focus of the G collimating lenses is aligned with the nominal imaging position of the G object pinholes. The G measured wavefronts are collimated by the G collimating lenses in the multi-channel Shack-Hartmann wavefront sensor. Each collimated wavefront is focused by the microlens array on the CMOS to form an image point array. The wavefront errors of the G viewpoints are obtained by analyzing the coordinates of the image point array of the G viewpoints. Based on the Z2~Z4 Zernike coefficients of the measured G wavefront errors, the calculation principle is the same as that of the multi-channel point diffraction interferometer sensor. The alignment position deviation (δx i ,δy i ,δz i ) and imaging position offset (Δx i ,Δy i)(I=1, i=1~G), where f in formula (3) is the focal length of the collimating lens in the multi-channel Shack-Hartmann wavefront sensor, and the alignment position deviation is the position deviation between the actual image point of each pinhole mark in the first group and the reference point (the focus of the collimating lens of the Shack-Hartmann wavefront sensor); the imaging position offset (Δx i ,Δy i )(I=2, i=1~G), until the imaging position offset (Δx i ,Δy i )(I=1~A, i=1~G).

[0038] The test steps using a multi-channel shearing interferometer sensor are as follows:

[0039] (1) Obtain the interference images of the first group G object surface test marks (I = 1, i = 1 to G)

[0040] a. Test the first group of G interference images in the first direction to obtain differential wavefront

[0041] Move the workpiece stage to align the nominal imaging positions of the G chessboard gratings with the G one-dimensional gratings in the first direction of the first group of the object plane gratings, and record the workpiece stage position [X I1 , Y I1 , Z I1 ](I=1). An incoherent light source uniformly illuminates G one-dimensional gratings in the first direction on the object surface. The one-dimensional grating modulates the spatial coherence of the light field and uniformly illuminates the pupil of the projection objective. The image plane checkerboard grating acts as a beam splitter to shift the diffraction orders relative to each other. After the zero-order diffraction light interferes with the other odd-numbered orders, G shearing interference images in the first direction are formed on the CCD. No interference occurs between the other diffraction orders. A phase shift algorithm is used to solve the first group of G interference images in the first direction to obtain the differential wavefront.

[0042] b. Test the first group of G interference images in the second direction to obtain the differential wavefront

[0043] Move the workpiece stage to align the G chessboard gratings with the nominal imaging position of the first group of G second direction gratings, and record the workpiece stage position at this time [X I2 , Y I2 , Z I2 ](I=1). G second-direction grating shearing interference images are formed on the CCD; a phase shift algorithm is used to solve the first group of G second-direction interference images to obtain a differential wavefront;

[0044] (2) Calculate the imaging position offset (Δx i ,Δyi )(I=1,i=1~G)

[0045] The wavefront reconstruction is used to obtain the wavefront aberration of the measured projection lens. The distortion can be obtained from the Zernike coefficients of Z2 to Z3. The alignment position deviation (δx i ,δy i )(I=1, i=1~G) can be calculated using formula (5):

[0046]

[0047] , alignment position deviation δx i ,δy i are the deviations between the positions of the G checkerboard gratings and the actual imaging positions of the first group of G object plane test marks, where P is the grating period, a2 and a3 are the Zernike coefficients of the Z2-Z3 terms in wavelength λ, s is the normalized shear amount, and the workpiece stage position [X I1 , Y I1 ](I=1) and [X I2 , Y I2 ](I=1), i.e. the actual imaging position of the first group G object plane test marks (x ri ,y ri ), calculate the imaging position offset (Δx i ,Δy i )(I=1,i=1~G);

[0048] (3) Calculate the imaging position offset (Δx i ,Δy i )(I=2~A,i=1~G)

[0049] Measure the imaging position offset (Δx i ,Δy i )(I=2~A, i=1~G), until the imaging position offset (Δx i ,Δy i )(I=1~A, i=1~G).

[0050] The imaging position offset (Δx i ,Δy i )(I=1~A, i=1~G) are fitted:

[0051]

[0052] , where T Ix ,θIx 、T Iy ,θ Iy The translation and rotation errors of the workpiece stage are respectively when the checkerboard grating is aligned with the G first-direction one-dimensional gratings and the second-direction one-dimensional gratings of the first group. N object plane test marks generate 2N equations, and the number of unknowns is 4A+3. When the number of equations is greater than or equal to the number of unknowns, that is,

[0053]

[0054] , fitting can be achieved. Since A>=1, G>=3.5, and since G is an integer, G>=4.

[0055] The beneficial effects of the present invention are:

[0056] 1. The accuracy of distortion detection is not affected by the positioning error of the workpiece stage;

[0057] 2. Fast measurement speed;

[0058] 3. The test steps are simple;

[0059] 4. Can simultaneously measure the wavefront aberration and distortion of the projection objective;

[0060] 5. The positioning error of the workpiece table can be measured. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a structural schematic diagram of a multi-channel objective lens distortion and magnification detection device according to the present invention;

[0062] Figure 2 Schematic diagram of the device for measuring the distortion and magnification of a projection objective lens based on a multi-channel point diffraction interferometer sensor in Example 1;

[0063] Figure 3 is a schematic diagram of the test mask in Example 1;

[0064] Figure 4 is a schematic diagram of the image plane test mark in Example 1;

[0065] Figure 5 Schematic diagram of a device for detecting projection lens distortion and magnification based on a multi-channel Shack-Hartmann sensor in Example 2;

[0066] Figure 6 is a schematic diagram of the mask in Example 2;

[0067] Figure 7 is a schematic diagram of the multi-channel Shack-Hartmann sensor in Example 2;

[0068] Figure 8 This is a schematic diagram of the device for measuring the distortion and magnification of the projection objective lens based on the multi-channel shearing interferometer wavefront sensor in Example 3. DETAILED DESCRIPTION

[0069] The following will further describe in detail the method for detecting the distortion and magnification of a multi-channel objective lens of the present invention in conjunction with a preferred embodiment.

[0070] Figure 1 The present invention provides a multi-channel objective lens distortion and magnification detection device, such as Figure 1 As shown, it includes an illumination system 100, a test mask 200, a mask stage 400, a projection objective lens 500, a multi-channel image plane sensor 600, and a workpiece stage 900. Among them, the illumination system 100 is used to provide a light source, the test mask 200 is mounted on the mask stage 400, and the lower surface of the test mask 200 is in contact with the upper surface of the mask stage 400. The lower surface of the test mask 200 includes A = 1 group, G = 3 object plane test marks 300, with a spacing of p, arranged along a straight line; the projection objective lens 500 is used to image the object plane test marks 300, there is a distance between the mask stage 300 and the projection objective lens 500, there is a distance between the projection objective lens 500 and the multi-channel image plane sensor 600, and the multi-channel image plane sensor 600 is mounted on the workpiece stage 9. 00, is used to simultaneously detect the imaging position offset between the actual image point and the nominal image point of multiple object plane test marks 300 after imaging by the projection objective lens 500. The multi-channel image plane sensor 600 includes three image plane test marks 700 conjugated with the object plane test mark 300 and one array detector 800. The image plane test marks 700 and the array detector 800 are arranged in the order of light transmission. The spacing p' of the image plane test marks 700 is p × M. The divergent light transmission distance between the array detector 800 and the plane where the image plane test marks 700 are located is q. The divergent light transmission distance is q. This ensures that light signals of different marks on the area array detector 800 of the multi-channel image sensor 600 do not interfere with each other; the workpiece stage 900 is used to carry the multi-channel image sensor 600.

[0071] Example 1

[0072] Figure 2 Schematic diagram of a device for detecting the distortion and magnification of a projection lens based on a multi-channel point diffraction interferometer sensor 600. The test mask 200 used is as follows: Figure 3 As shown, the test mask 200 includes 5×13 object plane test marks 300, and the object plane test marks 300 are double pinhole marks, including pinhole marks 301 and 302. The multi-channel point diffraction interferometer sensor 600 includes 1×13 image plane test marks 700, 1×13 collimating lenses 601, and a planar array detector 800, as shown in FIG. Figure 4As shown, the image plane test mark 700 includes a pinhole mark 701 and a light window 702. The front focal plane of the collimating lens 601 array is located on the plane where the image plane test mark 700 is located. The area array detector 800 is a CCD. The photosensitive surface of the area array detector 800 is located on the back focal plane of the collimating lens 601 array. The focal lengths of the 13 collimating lenses 601 are equal, and the focal length is f1. This ensures that the optical signals of different marks on the area array detector 800 of the multi-channel image sensor 600 do not interfere with each other.

[0073] The imaging position offset (Δx i ,Δy i )(I=1-5, i=1-13) are as follows:

[0074] The illumination system 100 is configured so that the light emitted by it generates 26 ideal spherical waves after passing through the first row of 13 double pinhole marks 300 on the test mask 200. The ideal spherical waves are imaged by the projection objective lens 500 to form 26 measured wavefronts. The workpiece stage 900 is moved. When the pinhole mark 701 in the 13 image plane test marks 700 of the multi-channel point diffraction interferometer sensor 600 is aligned with the nominal imaging position of the pinhole mark 302 in the first row of 13 double pinhole marks 300, the position of the workpiece stage 900 at this time is recorded. I , Y I , Z I ](I=1), one path (13) of the 26 measured wavefronts is diffracted by the 13 pinhole markers 702 to form an ideal spherical wave as the reference light, and one path (13) passes through the light window 702 as the measuring light. The 13 measuring light beams and the 13 reference light beams interfere with each other and are collimated by the 13 collimating lenses 601, resulting in 13 interference fringes on the CCD 800. The interference fringes are analyzed to obtain the wavefront errors of the 13 field points. Based on the Zernike coefficients of the 13 wavefront errors measured, the alignment position deviations (δx i ,δy i ,δz i ),

[0075]

[0076] , where Rsensor is the wavefront radius detected by CCD800, f is the focal length of the collimating lens 601 in the multi-channel point diffraction interferometer 600, and the alignment position deviation (δx i ,δy i ,δz i) is the position deviation between the actual image points of the first group of 13 pinhole marks 302 and the reference point (the focus of the collimating lens 601 of the point diffraction interferometer 600), and (δx i ,δy i ,δz i ), add the position of the workpiece table 900 [X I , Y I , Z I ](I=1), which is the actual imaging point position (x ri ,y ri , z ri ), calculate the nominal imaging position (x) of the pinhole mark 302 according to formula (2) i ,y i ):

[0077] (x i ,y i )=(xx i ,yy i )·M (2)

[0078] , then the imaging position offset (Δx i ,Δy i )(I=1, i=1-13):

[0079] (Δx i ,Δy i )=(x ri ,y ri )-(x i ,y i ) (4)

[0080] , change the setting of the illumination system 100 so that the light source it emits illuminates only the 13 double pinhole marks 300 in the second row, move the workpiece stage 900, and use the multi-channel point diffraction interferometer 600 to simultaneously image the 13 double pinhole marks 300 in the second row through the projection objective lens 500 to calculate the imaging position offset (Δx i ,Δy i )(I=2, i=1-13) to detect, and so on, until the imaging position offset (Δx i ,Δy i )(I=1~5,i=1~13); Fit the imaging position offset (Δx i ,Δy i )(I=1-5, i=1-13):

[0081]

[0082] , where the imaging position offsets in the x and y directions (Δx i ,Δy i ) are measured at the same time, so the rotation error θ1 of the workpiece stage 900 is the same. The equation group has 130 equations and 18 unknowns: 5 groups of positioning errors T of the workpiece stage 900 Ix 、T Iy ,θ I (I=1-5), the distortion of the projection lens 500 and the magnification error in the x and y directions: k1, M x 、M y , the unknown number can be obtained by fitting through the least square method.

[0083] Example 2

[0084] Figure 5 FIG. 6 is a schematic diagram of a device for detecting the distortion and magnification of a projection lens 500 based on a multi-channel Shack-Hartmann sensor 600. The test mask 200 used is as follows: Figure 6 As shown, the test mask 200 includes 1×3 object plane test marks 300, and the object plane test marks 300 are pinhole marks. The schematic diagram of the multi-channel Shack-Hartmann sensor 600 is shown in FIG. Figure 7 As shown, it includes 1×3 collimating lenses 601, a microlens array 602, a planar array detector 800, and three image plane test marks 700. The image plane test mark 700 is a light-transmitting window located at the front focal plane of the collimating lens array 601 and is conjugate to the object plane test mark 300. The planar array detector 800 is a CMOS. The microlens array 602 is located at the back focal plane of the collimating lens array 601. The photosensitive surface of the planar array detector 800 is located at the focal plane of the microlens array 602. The focal lengths of the three collimating lenses 601 are equal, that is, f2. This ensures that the optical signals of different labels on the area array detector 800 of the multi-channel Shack-Hartmann sensor 600 do not interfere with each other.

[0085] The imaging position offset (Δx i ,Δy i )(I=1, i=1-3) are as follows:

[0086] The light emitted by the illumination system 100 forms three ideal spherical waves after passing through the three pinhole marks 300. The ideal spherical waves are imaged by the projection objective lens 500 to form three measured wavefronts. The multi-channel Shack-Hartmann sensor 600 is moved until the three image plane test marks 700 are aligned with the nominal imaging positions of the three pinhole marks 300. The three measured wavefronts are collimated by three collimating lenses 601. Each collimated wavefront is focused by the microlens array 602 on the CMOS 800 to form an image point array. The wavefront errors of the three field of view points are obtained by analyzing the coordinates of the image point array of the three pinhole marks 300. Based on the Z2 to Z4 Zernike coefficients of the three measured wavefront errors, the alignment position deviation (δx i ,δy i ,δz i ) and imaging position offset (Δx i ,Δy i )(I=1, i=1-3), where f in formula (3) is the focal length of the collimating lens 601 in the multi-channel Shack-Hartmann wavefront sensor 600, and the alignment position deviation is the position deviation between the actual image point of each pinhole mark 300 and the reference point (the focus of the collimating lens 601 array of the Shack-Hartmann wavefront sensor 600); the imaging position offset (Δx i ,Δy i )(I=1,i=1~3) is fitted according to the following formula,

[0087]

[0088] , where the imaging position offsets in the x and y directions (Δx i ,Δy i ) are measured at the same time, so the rotation error θ1 of the workpiece stage 900 is the same. The fitting formula has 6 equations and 6 unknowns: the positioning error T of the workpiece stage 900 Ix 、T Iy ,θ I , the distortion of the projection lens 500 and the magnification error in the x and y directions: k1, M x 、M y , the unknown number can be obtained by fitting through the least square method.

[0089] Example 3

[0090] Figure 8Schematic diagram of a device for detecting the distortion and magnification of a projection objective lens 500 based on a multi-channel shearing interferometer sensor 600. The multi-channel shearing interferometer sensor 600 includes 1×4 image plane test marks 700 and an array detector 800. The test mask 200 includes 1×4 object plane test marks 300. The object plane test marks 300 include two orthogonal one-dimensional gratings 303 and 304, which are referred to as the first direction and the second direction. The one-dimensional gratings in the first and second directions are used to measure the imaging position offset in the x and y directions, respectively. The grating size and grating period in each direction are the same. The image plane test marks 700 are checkerboard gratings located conjugately with the object plane test marks 300. The array detector 800 is a CCD. Along the light transmission direction, the image plane test marks 700 and the array detector 800 are arranged in order. The distance between the plane where the image plane test marks 700 are located and the photosensitive surface of the array detector 800 is shown in FIG. This ensures that the optical signals of different marks on the area array detector 800 of the multi-channel shearing interferometer sensor 600 do not interfere with each other.

[0091] The steps for testing the imaging position offset using the multi-channel shearing interferometer wavefront sensor 600 are:

[0092] a. Test the interference images in the four first directions to obtain the differential wavefront

[0093] Move the workpiece stage to align the four chessboard gratings 700 with the nominal imaging position of the one-dimensional grating 303 in the first direction of the four object plane test marks 300, and record the workpiece stage position [X I1 , Y I1 , Z I1 An incoherent light source uniformly illuminates the one-dimensional grating 303 in four first directions on the object surface. The one-dimensional grating 300 modulates the spatial coherence of the light field and uniformly illuminates the pupil of the projection objective 500. The checkerboard grating 700 acts as a spectroscopic element to shift the diffraction orders relative to each other. After the zero-order diffraction light interferes with the other odd-numbered order diffraction lights, four shear interference images in the first directions are formed on the CCD 800. No interference occurs between the other diffraction orders. A phase shift algorithm is used to solve the interference images in the four first directions to obtain the differential wavefront.

[0094] b. Test the interference images in the four second directions to obtain the differential wavefront

[0095] Move the workpiece stage to align the four chessboard gratings with the nominal imaging positions of the four second direction gratings 304, and record the workpiece stage position [X I2 , Y I2 , Z I2 ]. Four shearing interference images in the second direction are formed on the CCD800; the phase shift algorithm is used to solve the interference images in the four second directions to obtain the differential wavefront;

[0096] c. Calculate the imaging position offset (Δx i ,Δy i )(I=1,i=1~4)

[0097] The wavefront reconstruction is used to obtain the wavefront aberration of the projecting lens 500 under test. The distortion can be obtained from the Zernike coefficients of Z2 to Z3. The alignment position deviation (δx i ,δy i )(I=1,i=1~4) can be calculated using formula (5):

[0098]

[0099] , alignment position deviation δx i ,δy i are the deviations between the positions of the four checkerboard gratings 700 and the actual imaging positions of the four object plane test marks 300, where P is the grating period, a2 and a3 are the Zernike coefficients of the Z2-Z3 terms in wavelength λ, s is the normalized shear amount, and the workpiece stage position [X I1 , Y I1 ](I=1) and [X I2 , Y I2 ](I=1), namely the actual imaging position (x ri ,y ri ), calculate the imaging position offset (Δx i ,Δy i )(I=1, i=1~4);

[0100] The imaging position offset (Δx i ,Δy i )(I=1, i=1~4) are fitted:

[0101]

[0102] Among them, T Ix 、T Iy are the translation errors of the workpiece stage 900 when aligning four checkerboard gratings 700 with four first-direction one-dimensional gratings 301 and second-direction one-dimensional gratings 302, θ Ix ,θ Iy The rotation errors of the workpiece stage 900 are respectively generated when four checkerboard gratings 700 are aligned with four first-direction one-dimensional gratings 301 and four second-direction one-dimensional gratings 302. The four object plane test marks 300 generate eight equations with seven unknowns, namely: Workpiece stage positioning error: T Ix ,θIx 、T Iy ,θ Iy And the distortion of the projection lens 500 and the magnification error in the x and y directions: k1, M x 、M y , the unknown number can be obtained by fitting through the least square method.

[0103] Matters not covered by the present invention are known technologies.

[0104] The above embodiments are intended only to illustrate the technical concepts and features of the present invention. Their purpose is to enable those skilled in the art to understand the contents of the present invention and implement them accordingly. They are not intended to limit the scope of protection of the present invention. Any equivalent changes or modifications made in accordance with the spirit of the present invention are intended to be covered by the scope of protection of the present invention.

Claims

1. A multi-channel objective lens distortion and magnification detection device, comprising an illumination system (100), a mask stage (400), a test mask (200), a projection objective (500), a workpiece stage (900), and a sensor, wherein the illumination system (100) is used to provide a light source; the mask stage (400) is used to carry the test mask (200); and the device is characterized in that: The test mask (200) is engraved with A groups, each group having G object plane test marks, a total of N, where A>=1, G>=3, and N>=3. The G object plane test marks in each group have the same spacing p and spatial distribution, and are arranged along a straight line, a broken line, or an array according to the spacing p. The spacing between different groups is greater than or equal to p. The object plane test marks (300) are all located within the object field of view of the projection objective lens (500) to be tested. The projection objective lens (500) is used to image the object surface test mark (300); the sensor is a multi-channel image plane sensor (600) used to simultaneously detect the imaging position offset between the actual image points and the nominal image points of multiple object surface test marks after imaging through the projection objective lens; The workpiece stage (900) moves during the detection process so that the image plane test mark of the multi-channel image plane sensor (600) is aligned with the nominal imaging position of the corresponding object plane test mark; The multi-channel image plane sensor (600) comprises a group of G image plane test marks in a conjugate relationship with a group of G object plane test marks and an array detector (800), wherein the spacing distance p' of each image plane test mark is equal to M×p, wherein M is the nominal zoom magnification of the projection objective lens (500); the image plane test marks and the array detector are sequentially arranged along the light transmission direction, and the divergent light transmission distance q between the photosensitive surface of the array detector and the plane where the image plane test marks are located is This ensures that light signals of different marks on the area array detector of the multi-channel image sensor do not interfere with each other.

2. The multi-channel objective lens distortion and magnification detection device according to claim 1, characterized in that: Each object plane test mark (300) includes two pinhole marks (301, 302); the multi-channel image plane sensor (600) is a multi-channel point diffraction interferometer sensor, including G image plane test marks (700), a collimating lens array, and an array detector (800); the image plane test mark (700) includes a pinhole mark (701) and a light window (702); the collimating lens array includes G collimating lenses (601); the front focal plane of the collimating lens array is located at the plane where the image plane test mark (700) is located; the array detector (800) is a CCD; the photosensitive surface of the array detector (800) is located at the back focal plane of the collimating lens array; the focal lengths of the G collimating lenses (601) are equal, and the focal length is f1. This ensures that light signals of different marks on the area array detector (800) of the multi-channel image plane sensor (600) do not interfere with each other.

3. The multi-channel objective lens distortion and magnification detection device according to claim 1, characterized in that: Each object plane test mark (300) is a pinhole mark; the multi-channel image plane sensor (600) is a multi-channel Shack-Hartmann sensor, comprising the image plane test mark (700), a collimating lens array, a microlens array (602) and an area array detector (800); the image plane test mark (700) is a light-transmitting window; the collimating lens array comprises G collimating lenses (601); the front focal plane of the collimating lens array is located on a plane that is conjugate to the object plane test mark (300); the area array detector (800) is a CMOS; the microlens array (602) is located on the back focal plane of the collimating lens array; and the photosensitive surface of the area array detector (800) is located on the focal plane of the microlens array (602); the focal lengths of the G collimating lenses (601) are equal, i.e., f2. This ensures that light signals of different marks on the area array detector (800) of the multi-channel image plane sensor (600) do not interfere with each other.

4. The multi-channel objective lens distortion and magnification detection device according to claim 1, characterized in that: The object plane test mark (300) comprises one-dimensional gratings (303, 304) in two orthogonal directions, which are referred to as a first direction and a second direction, and the grating size and grating period in each direction are the same; the multi-channel image plane sensor (600) is a multi-channel shearing interferometer sensor, comprising an image plane test mark (700) and an array detector (800); the image plane test mark (700) is a checkerboard grating and is located on a plane that is conjugate to the object plane test mark (300); the array detector (800) is a CCD, and the image plane test mark (700) and the array detector (800) are located in sequence along the light transmission direction; the divergent light transmission distance between the photosensitive surface of the array detector (800) and the plane where the image plane test mark (700) is located is This ensures that light signals of different marks on the area array detector (800) of the multi-channel image plane sensor (600) do not interfere with each other.

5. A method for detecting distortion and magnification of a multi-channel objective lens, characterized in that: The multi-channel objective lens distortion and magnification detection device according to any one of claims 1 to 4 is used for testing, characterized in that it comprises: Step 1: Mount the test mask (200) on a mask stage (400), and move the object plane test mark (300) into the field of view of the projection objective lens (500); Step 2: Mounting the multi-channel image plane sensor (600) on a workpiece stage (900) and moving it to a plane that is conjugate with the object plane test mark (300); Step 3: Using the illumination system (100), the object plane test mark (300) is imaged through the projection objective lens (500), so that the image plane test mark of the multi-channel image plane sensor is aligned with the nominal imaging position of the corresponding object plane test mark, and using the multi-channel image plane sensor (600) to simultaneously image the first group G of the object plane test marks (300) through the projection objective lens (500) to obtain the imaging position offset (Δx i ,Δy i ) is tested, and after the imaging position offset test of the group G channels is completed, the position of the test field point is changed until the imaging position offset (Δx i ,Δy i ); Step 4: (Δx i ,Δy i ) is used to perform least square fitting to calculate the distortion and magnification error of the projection objective lens. The fitting equation is: Among them, T Ix 、T Iy are respectively the translation errors of the workpiece stage (900) when the multi-channel image plane sensor (600) is used to detect the imaging position offsets of the object plane test marks (300) in the x-direction and the y-direction of the Ith (I=1, 2...A)th group, θ Ix ,θ Iy are respectively the rotation errors of the workpiece stage (900) when the multi-channel image plane sensor (600) is used to detect the imaging position offsets of the object plane test marks (300) in the x-direction and the y-direction of the I (I=1, 2...A) group. x 、M y is the magnification of the projection objective lens (500) in the x-direction and the y-direction, is the distance from the nominal imaging position of the i-th object plane test mark (300) to the center of the image plane coordinate system, k1 is the third-order distortion coefficient of the projection objective lens (500), r x 、r y is the fitting residual, (x i ,y i ) is the nominal imaging position of the i-th (i=1, 2...G) object plane test mark (300) in the I-th group, and the formula is as follows: (x i ,y i )=(xx i ,yy i )·M (2) Among them, (xx i ,yy i ) is the position of the object plane test mark (300) in the object plane coordinate system.

6. The method for detecting distortion and magnification of a multi-channel objective lens according to claim 5, wherein: In step 3, the position of the test field point is changed, specifically: When the test mask (200) includes only one group of G object surface test marks (300), the mask stage (400) is moved so that the light emitted by the illumination system (100) forms an image of the G object surface test marks (300) through the projection objective lens (500), and the workpiece stage (900) is moved so that the multi-channel image plane sensor (600) simultaneously forms an image of the G object surface test marks (300) through the projection objective lens (500). i ,Δy i ) for detection; moving the mask stage (400) to change the position of the field point where the object plane test mark (300) is located, and moving the workpiece stage (900) so that the multi-channel image plane sensor (600) once again simultaneously images the G object plane test marks (300) through the projection objective lens (500) to obtain the imaging position offset (Δx i ,Δy i ) for detection; and so on, until the imaging position offset (Δx i ,Δy i ); When the test mask (200) includes A groups, each group of G object surface test marks (300), the illumination system (100) is set so that the illumination light it emits illuminates only the first group of G object surface test marks (300) and images them through the projection objective lens (500), and the workpiece stage (900) is moved so that the multi-channel image plane sensor (600) simultaneously images the first group of G object surface test marks (300) through the projection objective lens (500). The imaging position offset (Δx i ,Δy i ) for detection; changing the setting of the illumination system (100) so that the illumination light emitted by it illuminates only the second group G of the object surface test marks (300); moving the workpiece stage (900) so that the multi-channel image plane sensor (600) simultaneously images the second group G of the object surface test marks (300) through the projection objective lens (500) to determine the imaging position offset (Δx i ,Δy i ) for detection; and so on, until the imaging position offset (Δx i ,Δy i ).

7. The method for detecting distortion and magnification of a multi-channel objective lens according to claim 5 when referring to claim 2, characterized in that: The object plane test mark (300) is a double pinhole mark, the multi-channel image plane sensor (600) is a multi-channel point diffraction interferometer sensor, and the method for detecting the imaging position offset in step 3 is: The illumination light source (100) generates 2G ideal spherical waves after passing through the first group of G double pinhole marks on the test mask (200). The ideal spherical waves are imaged by the projection objective lens (500) to form 2G measured wavefronts. The workpiece stage (900) is moved. When the G image pinhole marks (701) of the multi-channel point diffraction interferometer (600) are aligned with the nominal imaging positions of the G object surface pinholes (302), the position [x1, y1, z1] of the workpiece stage (900) at this time is recorded. The path of the 2G measured wavefronts is recorded. G ideal spherical waves are formed after being diffracted by G image plane pinhole marks (701) and serve as reference light. G waves pass through the optical window (702) and serve as measurement light. The G measurement lights and the G reference lights interfere with each other and are collimated by G collimating lenses (601). G interference fringes are presented on the CCD (800). The interference fringes are analyzed to obtain the wavefront errors of the first group of G field points. Based on the Z2 to Z4 Zernike coefficients of the measured G wavefront errors, the alignment position deviation (δx i ,δy i ,δz i ), Wherein, Rsensor is the wavefront radius detected by the CCD (800), f is the focal length of the collimating lens (601) in the multi-channel point diffraction interferometer sensor (600), (δx i ,δy i ,δz i ) is the positional deviation between the actual image point of each pinhole mark (302) in the first group and the reference point, the reference point being the focus of the collimating lens (601) of the point diffraction interferometer sensor, and (δx i ,δy i ,δz i ), and then add the wafer stage position [X1, Y1, Z1], which is the actual imaging point position (x ri ,y ri , z ri ), the nominal imaging position (x) of the object plane pinhole mark (300) is calculated by formula (2): i ,y i ), then the imaging position offset (Δx i ,Δy i The calculation formula of (Δx )(I=1,i=1~G) is i ,Δy i )=(x ri ,y ri )-(x i ,y i ) (4) Move the field of view and measure the imaging position offset (Δx i ,Δy i )(I=2, i=1-G), until the imaging position offset (Δx i ,Δy i )(I=1~A).

8. The method for detecting distortion and magnification of a multi-channel objective lens according to claim 5 when referring to claim 3, characterized in that: The object plane test mark (300) is a pinhole mark, the multi-channel image plane sensor (600) is a multi-channel Shack-Hartmann sensor, and the method for detecting the imaging position offset in step 3 is: Light emitted by the illumination system (100) forms G ideal spherical waves after passing through a first group of G object surface test marks (300). The ideal spherical waves are imaged by the projection objective lens (500) to form G measured wavefronts. The multi-channel Shack-Hartmann sensor is moved until the front focal points of the G collimating lenses (601) are aligned with the nominal imaging positions of the G object surface test marks (300). The G measured wavefronts are collimated by the G collimating lenses (601) in the multi-channel Shack-Hartmann wavefront sensor. Each collimated wavefront is focused by a microlens array (602) on a CMOS to form an image point array. The wavefront errors of the G viewpoints are obtained by analyzing the image point array coordinates of the G viewpoints. Based on the Zernike coefficients of the measured G wavefront errors, the following formula is used: (x i ,y i )=(xx i ,yy i )·M (5) Among them, (xx i ,yy i ) is the position of the object plane test mark (300) in the object plane coordinate system; Wherein, Rsensor is the wavefront radius detected by the CMOS (800), and f is the focal length of the collimating lens (601) in the multi-channel Shack-Hartmann wavefront sensor; (Δx i ,Δy i )=(x ri ,y ri )-(x i ,y i ) (7) Calculate the alignment position deviation (δx i ,δy i ,δz i ) and imaging position offset (Δx i ,Δy i )(I=1, i=1~G), the alignment position deviation is the position deviation between the actual image point of each pinhole mark of the first group and the reference point, the reference point is the focus of the collimating lens (601) of the Shack-Hartmann wavefront sensor; the imaging position offset (Δx i ,Δy i )(I=2, i=1-G), until the imaging position offset (Δx i ,Δy i )(I=1~A, i=1~G).

9. The method for detecting distortion and magnification of a multi-channel objective lens according to claim 5 when referring to claim 4, characterized in that: The object plane test mark (300) is a one-dimensional grating, the multi-channel image plane sensor (600) is a multi-channel shearing interferometer sensor, and the method for detecting the imaging position offset in step 3 is: Move the workpiece stage to align the G chessboard gratings (700) with the nominal imaging positions of the G first-direction one-dimensional gratings (301) in the first group of the object surface test marks (300), and record the position of the workpiece stage (900) at this time [X I1 , Y I1 ] (I=1), an incoherent light source uniformly illuminates the G first-direction one-dimensional grating (301) arrays on the object surface, the one-dimensional grating (300) modulates the spatial coherence of the light field, and causes the pupil of the projection objective lens (500) to be uniformly illuminated; the chessboard grating (700) on the image surface acts as a light splitting element to shift and misalign the diffraction orders relative to each other, and after interference between the 0th order and the other odd-numbered order diffraction lights, G first-direction shear interference images are formed on the CCD (800), and no interference occurs between the other diffraction orders; the workpiece stage (900) is moved to align the G chessboard gratings (700) with the nominal imaging positions of the first group of G second-direction gratings (302), and the workpiece stage (900) is positioned [X I2 , Y I2 ](I=1), forming G shearing interference images of the second direction grating (302) on the CCD (800); using a phase shift algorithm to solve the first group of G first direction and G second direction interference images to obtain a differential wavefront, reconstructing the wavefront to obtain the wave aberration of the measured projection objective lens (500), and calculating the distortion from the Z2-Z3 Zernike coefficients. The alignment position deviation based on the shearing interference is calculated using the following formula: The alignment position deviation is the deviation between the position of the G chessboard gratings (700) and the actual imaging position of the first group of the G object plane test marks (300), wherein P is the grating period, a2 and a3 are the Zernike coefficients of the Z2-Z3 items, the unit of which is wavelength λ, s is the normalized shear amount, and the position of the workpiece stage (900) [X I1 , Y I1 ] (I=1) and [Y I2 , Y I2 ](I=1), i.e. the actual imaging position (x ri ,y ri , z ri ), according to the formula: (Δx i ,Δy i )=(x ri ,y ri )-(x i ,y i ) (9) Calculate the imaging position offset (Δx i ,Δy i )(I=1, i=1~G); measuring the imaging position offset (Δx i ,Δy i )(I=2, i=1-G), until the imaging position offset (Δx i ,Δy i )(I=1~A, i=1~G).

Images