A method for horizontal calibration error monitoring and compensation of an LDI exposure machine

Abstract

The application discloses a kind of for LDI exposure machine horizontal direction calibration error monitoring and compensation method, S1: before exposure to material plate, to platform, camera and exposure lens are calibrated horizontally;S2: each measuring point on the figure coordinate system where material plate is located is converted to platform coordinate system, and the theoretical position coordinates of each measuring point are obtained;After completing horizontal direction calibration, exposure is carried out, by the difference between the theoretical position and actual position of each measuring point is fitted, and the expansion, rotation error of platform in exposure position and the position calibration error of reference lens and reference camera are calculated and compensated, the calibration error of exposure lens magnification, angle and exposure lens position relationship is monitored and compensated, the grid error of platform in exposure position is monitored and compensated, the error generated by each subsystem in horizontal direction can be decomposed and compensated, finally ensure the accuracy of overlay.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of semiconductor lithography equipment, and particularly relates to a method for horizontal calibration error monitoring and compensation of an LDI exposure machine. BACKGROUND

[0002] Overlay is an important index for horizontal acceptance of an LDI exposure machine. An exposure process of the exposure machine is as follows: a double-table direct writing exposure machine (publication number CN211741826U) disclosed in a Chinese patent is provided with an exposure mechanism, a moving workpiece table, a positioning system and other components. Due to the influence of horizontal calibration error factors, a certain deviation exists between the center positions of a circle and a ring during the overlay test, and the large overlay error will reduce the yield in subsequent production.

[0003] Moreover, the horizontal error value is composed of error coupling of multiple subsystems. The existing scheme cannot analyze the overlay existing data, decompose and compensate the error generated by each subsystem without adding additional sensors, finally guaranteeing the accuracy of the overlay, and cannot intuitively obtain the contribution of each subsystem to the error, so as to compensate accordingly.

[0004] Therefore, the present application provides a method for horizontal calibration error monitoring and compensation of an LDI exposure machine to solve the problems in the background art. SUMMARY

[0005] The present application aims to provide a method for horizontal calibration error monitoring and compensation of an LDI exposure machine to solve the problems in the background art.

[0006] To achieve the above-mentioned purpose, the present application provides the following technical scheme:

[0007] S1: Before exposing the material plate, calibrate the platform, the camera and the exposure lens horizontally.

[0008] S2: Convert each measurement point on the material plate in the graphic coordinate system to the platform coordinate system to obtain the theoretical position coordinates of each measurement point.

[0009] S3: Use the exposure lens to expose the ring and the circle in each measurement point on the material plate respectively, use the camera to capture the deviation of the center of the ring and the circle, and use the camera to obtain the actual position coordinates of the ring in the platform coordinate system, and calculate the difference between the actual position and the theoretical position of each measurement point.

[0010] S4: Fit the difference between the theoretical position and the actual position of each measurement point, calculate and compensate the expansion and rotation error of the platform at the exposure position, and the position calibration error of the reference lens and the reference camera, to obtain the first residual value.

[0011] S5: calculating and compensating the calibration error of the exposure lens magnification, angle calibration and exposure lens position relationship based on the first residual value, to obtain a second residual value;

[0012] S6: calculating and compensating the grid error of the platform at the exposure position based on the second residual value.

[0013] As a further scheme of the present application: in step S2, the affine transformation model is used to convert the coordinates of each measurement point on the to-be-exposed pattern before exposure, to obtain the positions of each measurement point of the to-be-exposed pattern in the platform coordinate system under the theoretical state.

[0014] As a further scheme of the present application: step S3 specifically includes:

[0015] S3-1: the deviation of the camera from the center of the circle is the threading and scribing error, the camera is divided according to the position relationship of the measurement points in the platform coordinate system, to obtain the scanning period, exposure lens and strip label to which each measurement point belongs;

[0016] S3-2: subtracting the platform coordinates of each measurement point actually captured by the camera from the platform coordinates of the theoretical position of each measurement point, to obtain the difference between the theoretical position and the actual position of each measurement point.

[0017] As a further scheme of the present application: step S4 specifically includes:

[0018] S4-1: using the affine transformation model to fit the difference between the theoretical and actual positions of each measurement point by the least square method, and compensating the fitting coefficients to the affine transformation model respectively;

[0019] S4-2: obtaining the first residual value after removing the expansion and rotation error of the platform at the exposure position, and the calibration error of the reference lens and the reference camera position.

[0020] As a further scheme of the present application: step S5 specifically includes:

[0021] S5-1: according to the position relationship of the measurement points in the exposure lens coordinate system, the first residual value is fitted by the least square method according to the different scanning periods, exposure lenses and strip labels, to obtain the lens X direction magnification calibration error, lens angle calibration error and lens position relationship calibration error of each exposure lens;

[0022] S5-2: compensating the lens magnification calibration error, lens angle calibration error and lens position relationship calibration error to the lens magnification and angle calibration parameters and the position relationship between the lenses respectively;

[0023] S5-3: Obtain the second residual value after removing the lens magnification, angle calibration and exposure lens position relationship calibration error of the exposure lens.

[0024] As a further scheme of the present application, step S6 specifically comprises:

[0025] S6-1: According to the actual platform coordinates of each measurement point, the coordinates of the platform X and Y when the exposure measurement point is calculated, the second residual value after removing the lens magnification, angle calibration and exposure lens position relationship calibration error of the exposure lens is one-to-one corresponding to the platform position, and the grid error of the measurement points at similar positions of the platform is averaged to represent the effective value of the grid error of the platform when the exposure measurement point is calculated;

[0026] S6-2: The effective value of the grid data of each position of the platform is compensated into the platform grid error at the exposure position, and the remaining positions in the grid are filled in the form of linear interpolation.

[0027] As a further scheme of the present application, it further comprises S7: after compensating the error amount in steps S4-S6, re-exposing the plate to verify whether the compensation result is effective, and judging the convergence according to the compensation result;

[0028] S8: Calculate and compensate the calibration residual error of the alignment system.

[0029] As a further scheme of the present application, in step S7, verifying whether the compensation result is effective comprises: platform expansion and contraction error, reference lens and reference camera position relationship error, exposure lens magnification and angle error, exposure lens position relationship, and platform grid error at the exposure position.

[0030] As a further scheme of the present application, step S8 specifically comprises:

[0031] S8-1: First, expose the circular ring pattern, then select four points in the circular ring layer as alignment points to complete alignment, then expose the circle, and finally use the alignment camera to measure the center deviation of the circular ring and the circle. The error between the circular ring and the circle center is caused by the calibration error of the alignment system.

[0032] S8-2: Fit each point on the material plate to obtain the calibration residual error of the alignment system, compensate the calibration residual error into the alignment system, and add the compensation parameter of the calibration residual error of the alignment system when using the alignment function for subsequent exposure, so as to reduce the overlay deviation between the circular ring and the circle center, and optimize the single machine overlay performance of the machine.

[0033] Compared with the prior art, the present application has the following advantages:

[0034] After the exposure is completed, the expansion and rotation errors of the platform at the exposure position, the position calibration errors of the reference lens and the reference camera, the calibration errors of the exposure lens magnification, angle and position relationship are monitored and compensated by fitting the difference between the theoretical position and the actual position of each measuring point, the grid error of the platform at the exposure position is monitored and compensated, the errors generated by each subsystem in the horizontal direction can be decomposed and compensated, and finally the accuracy of the overlay is ensured. BRIEF DESCRIPTION OF DRAWINGS

[0035] Figure 1 It is an embodiment schematic diagram of a method for LDI exposure machine horizontal calibration error monitoring and compensation.

[0036] Figure 2 It is an exposure pattern schematic diagram in a method for LDI exposure machine horizontal calibration error monitoring and compensation.

[0037] Figure 3 It is an exposure lens and motion platform schematic diagram in a method for LDI exposure machine horizontal calibration error monitoring and compensation.

[0038] Figure 4 It is an overlay test flowchart in a method for LDI exposure machine horizontal calibration error monitoring and compensation. DETAILED DESCRIPTION

[0039] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0040] As shown in Figure 1 The present embodiment provides a method for LDI exposure machine horizontal calibration error monitoring and compensation, comprising the following steps:

[0041] S1: Before exposing the material plate, calibrate the platform, camera and exposure lens horizontally.

[0042] As shown in Figures 2-3 S2: Convert each measuring point on the pattern coordinate system of the material plate to the platform coordinate system to obtain the theoretical position coordinates of each measuring point.

[0043] Due to the influence of the motion error of the platform at the exposure position and the calibration error of the exposure lens, the exposed pattern deviates from the theoretical value. Before exposure, the affine transformation model is used to convert the coordinates of each measurement point on the to-be-exposed pattern, to obtain the positions of each measurement point of the to-be-exposed pattern in the platform coordinate system under the theoretical state. The matrix includes but is not limited to:

[0044]

[0045] Wherein, mx, my represent the expansion and contraction of the platform in the X, Y directions of the exposure position, rx, ry represent the rotation of the platform at the exposure position; tx, ty represent the overall translation of the platform at the exposure position, the positions of each measurement point on the exposure pattern in the pattern coordinate system are subjected to affine transformation model Matrix 1x and Matrix 1y The position coordinate set Xl, Yl of each measurement point of the exposure pattern in the platform coordinate system under the theoretical state can be obtained, wherein Xl = {Xl1, Xl2…Xl n}, Yl = {Yl1, Yl2…Yl n}

[0046] S3: Use the exposure lens to expose the ring and the circle in each measurement point on the material plate respectively, use the camera to capture the deviation of the ring and the circle center, and the camera obtains the actual position coordinates of the ring in the platform coordinate system, and calculates the difference between the actual position and the theoretical position of each measurement point.

[0047] S3-1: Without using the camera for alignment, use the exposure lens to expose the ring and the circle in each measurement point on the material plate respectively, use the camera to capture the deviation errx1, erry1 of the ring and the circle center, and the camera obtains the actual position coordinate set Xr, Yr of the ring in the platform coordinate system; the deviation errx1, erry1 of the ring and the circle center obtained by the camera is the threading and scribing error, and the camera is a CCD alignment camera; the threading and scribing error represents the repeatability performance of the platform in a short time, if the threading and scribing error exceeds 3um, it needs to be checked for abnormality.

[0048] The camera divides according to the position relationship of the measurement points in the platform coordinate system, obtains the scanning period, the exposure lens and the strip label to which each measurement point belongs.

[0049] S3-2: The positions Xl, Yl of each measurement point of the exposure pattern in the platform coordinate system under the theoretical state minus the actual position coordinates Xr, Yr of the ring in the platform coordinate system obtained by the camera:

[0050] ① Xl-Xr = Errx1

[0051] ② Yl-Yr = Erry1

[0052] The difference set Errx1, Erry1 between the theoretical position and the actual position of each measuring point is obtained, wherein Errx1 = {Errx11, Errx12,... Errx1 n}, and Erry1 = {Erry11, Erry12,... Erry1 n}.

[0053] S4: fitting the difference between the theoretical position and the actual position of each measuring point, calculating and compensating the expansion and contraction, rotation error of the platform at the exposure position, and the position calibration error of the reference lens and the reference camera, to obtain the first residual value, specifically including:

[0054] S4-1: using an affine transformation model to fit the difference between the theoretical and actual positions of each measuring point Errx1, Erry1 by the least square method: in the formula, two models are used to fit Errx1 and Err y1 respectively with coordinates, according to the principle of least square method, the least square method finds the best function matching of data by minimizing the sum of squares of errors, so that the sum of squares of errors between the data obtained and the actual data is minimized, but the model cannot express the distribution of Errx1 and Erry1 errors by 100% through the least square method fitting, so there will be the first residual value Errx2 and Erry2.

[0055]

[0056] xl, Yl represent the position coordinate set of each measuring point of the exposure pattern in the platform coordinate system under the theoretical state, Xl1, Xl2,... Xl n are components of the Xl set, Yl1, Yl2,... Yl n are components of the Yl set, Errx11, Errx12,... Errx1 n are components of the Errx1 set, Erry11, Erry12,... Erry1 n are components of the Erry1 set, Mx1, My1 represent the difference between the expansion and contraction of the platform in the X and Y directions at the exposure position and the theory, Rx1, Ry1 represent the difference between the rotation of the platform at the exposure position and the theory, Tx1, Ty1 represent the difference between the overall translation of the platform at the exposure position and the theory.

[0057] Then the fitted coefficients are named as Matrix 11x and Matrix 11y , and the fitting coefficients are compensated to the affine transformation model Matrix 1x and Matrix 1y ​​, used to remove the expansion, contraction, and rotation errors of the platform at the exposure position, as well as the calibration errors of the reference lens and reference camera positions, namely:

[0058]

[0059] S4-2: After removing the expansion and contraction errors and rotation errors of the platform at the exposure position and the calibration errors of the reference lens and reference camera, the residual values ​​are set as the first residual values ​​Errx2 and Erry2. According to the principle of least squares fitting, the model cannot fully express the error distribution through least squares fitting, so the first residual values ​​Errx2 and Erry2 will exist:

[0060]

[0061] S5: After removing the expansion, contraction, and rotation errors of the platform at the exposure position and the calibration errors of the reference lens and reference camera positions, calculate and compensate for the calibration errors of the exposure lens magnification, angle calibration, and exposure lens position relationship. Specifically, this includes:

[0062] S5-1: According to the positional relationship of the measurement points in the exposure lens coordinate system, the first residual values ​​Errx2 and Erry2 are respectively fitted with a first-order linear model according to the scanning cycle, exposure lens, and strip number, and the least squares method is used to obtain the lens X-direction magnification calibration error, lens angle calibration error, and lens position relationship calibration error of each exposure lens: The two models Fit Errx2 and Erry2 to the coordinates respectively. According to the principle of least squares, the model cannot fully express the distribution of Errx2 and Erry2 errors through least squares fitting, so there will be second residual values ​​Errx3 and Erry3.

[0063]

[0064] Among them, x is the position information of each measurement point in the exposure lens coordinate system, x1, x2, ..., x n is a component of the x set, and is obtained by inferring the position of each measurement point in the lens coordinate system through the coordinate relationship of each measurement point in the platform coordinate system and the scanning cycle, exposure lens, and strip label information; Mx(len) is the lens magnification calibration error; Rx(len) is the lens angle calibration error; Tx2, Ty2 are the calibration errors of the position relationship of each lens, Errx21, Errx22, ... Errx2 n It is a component of Errx2 set, Erry21, Erry22, ... Erry2 n It is part of the Erry2 collection.

[0065] S5-2: Compensate the lens magnification calibration error, the lens angle calibration error and the lens position relationship calibration error into the lens magnification and angle calibration parameters and the lens position relationship respectively;

[0066] S5-3: The first residual value Errx2, Erry2 is the residual value after the exposure lens magnification, angle calibration and exposure lens position relationship calibration error is removed, and the second residual value Errx3, Erry3 is set. According to the principle of least squares method, the least squares method finds the best function matching of data by minimizing the sum of squares of errors, so that the sum of squares of errors between the data obtained and the actual data is minimized. However, the model cannot express the distribution of Errx2, Erry2 error by 100% through the least squares fitting, so there will be a second residual value Errx3, Erry3:

[0067]

[0068] S6: After removing the lens magnification, angle calibration and exposure lens position relationship calibration error of the exposure lens, calculate and compensate the grid error of the platform at the exposure position, which specifically includes:

[0069] S6-1: Since the first-order and zero-order error amount of the platform at the exposure position is removed in step S4, but the remaining high-order error amount is not compensated, according to the actual platform coordinates of each measurement point, the coordinates of the platform X and Y are calculated when the exposure measurement point is calculated. The second residual value Errx3, Erry3 of each measurement point is one-to-one corresponding to the platform position, and the grid error of the measurement point at the similar position of the platform is averaged to represent the effective value of the grid error of the platform position when the exposure measurement point is calculated. For example, the grid error of the measurement point with a platform position distance within 100um is averaged, and the calculation equation is:

[0070] Err pos1 = (err pos1 + err pos2 + err pos3 + … + err posn ) / n

[0071] Wherein, Err pos1 represents the effective value of the platform grid when the exposure measurement point is calculated, err pos1 , err pos2 , err pos3 , …, err posn represents the grid value at the similar position of the platform position when the exposure measurement point is calculated. Thus, a group of grid error values can be obtained in the X and Y directions of the platform.

[0072] S6-2: The effective value of the grid data at each position of the platform is compensated to the platform grid error at the exposure position, so that the platform can move accurately to the theoretical position. The remaining positions in the grid are filled using linear interpolation.

[0073] S7: After compensating for the errors in steps S4-S6, re-expose the plate to verify whether the compensation results are effective, and judge the convergence based on the compensation results: if the platform expansion, contraction, and rotation errors converge to 4E-07, and the positional relationship error between the reference lens and the reference camera converges to within 0.4um, then step S4 is considered to be fully compensated; if the exposure lens magnification and angle errors converge to 3E-05, and the positional relationship between the exposure lenses converges to within 0.3um, then step S5 is considered to be fully compensated; if the average values ​​of the X and Y grid errors of the platform at the exposure position converge to within 0.3um, then step S6 is considered to be fully compensated, otherwise it is necessary to continue iterating or look for the cause of the abnormality.

[0074] S8: After verifying that the compensation is effective and converged, the plate is replaced again to expose the ring and circle on the material plate, and the calibration residual of the alignment system is calculated and compensated. Specifically, it includes:

[0075] S8-1: First expose the ring pattern, such as Figure 4 Ring 1; then select four points in the ring layer as alignment points to complete the alignment, such as Figure 4 Ring 2; then expose the circle, such as Figure 4 Circle 3; Finally, the alignment camera is used to measure the X / Y center deviations errx2 and erry2 of the ring and circle. The center errors of the ring and circle are caused by the calibration error of the alignment system.

[0076] S8-2: Fitting each point on the material surface:

[0077]

[0078] Get the calibration residual Matrix of the alignment system 2x 、Matrix 2y :

[0079]

[0080] Where X2 and Y2 represent the theoretical position of the ring in the platform coordinate system, X21, X21,..., X2 n is a component of the set X2, Y21, Y21, ..., Y2 n It is a component of the Y2 set; errx21, errx22, ..., errx2 n is a component of the errx2 set, erry21, erry22, ..., erry2 nMx2, My2 represent the difference between the expansion and contraction of the alignment system X, Y direction and the theory; Rx2, Ry2 represent the difference between the rotation of the exposure site of the alignment system and the theory; Tx2, Ty2 represent the difference between the translation of the whole alignment system and the theory;

[0081] The calibration residual compensation is compensated into the alignment system, and the calibration residual compensation parameter of the alignment system is added when the exposure is performed by using the alignment function subsequently, so as to reduce the overlay deviation between the circle and the center of the circle, and to optimize the single machine overlay performance of the machine.

[0082] The above merely describes the preferred embodiments of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can make equivalent replacements or changes according to the technical scheme and the inventive concept of the present application within the technical range disclosed by the present application, and all of them should be covered within the protection scope of the present application.

Claims

1. A method for LDI exposure machine horizontal calibration error monitoring and compensation, characterized in that, The method comprises the following steps: S1: before exposing the material plate, calibrate the platform, camera and exposure lens horizontally; S2: convert each measuring point on the graphic coordinate system where the material plate is located to the platform coordinate system to obtain the theoretical position coordinates of each measuring point; S3: use the exposure lens to expose each measuring point on the material plate, use the camera to capture the deviation of the circle ring and the circle center, and use the camera to obtain the actual position coordinates of the circle ring in the platform coordinate system, and calculate the difference between the actual position and the theoretical position of each measuring point; S4: fit the difference between the theoretical position and the actual position of each measuring point, calculate and compensate the expansion, rotation error of the platform at the exposure position, and the position calibration error of the reference lens and the reference camera, to obtain a first residual value; S5: based on the first residual value, calculate and compensate the magnification, angle calibration of the exposure lens and the calibration error of the position relationship of the exposure lens, to obtain a second residual value; S6: based on the second residual value, calculate and compensate the grid error of the platform at the exposure position.

2. The method for horizontal calibration error monitoring and compensation of an LDI exposure machine according to claim 1, wherein, In step S2, affine transformation model is used to convert the coordinates of each measuring point on the to-be-exposed graphic before exposure to obtain the position of each measuring point of the to-be-exposed graphic in the platform coordinate system under the theoretical state.

3. The method for horizontal calibration error monitoring and compensation of an LDI exposure machine according to claim 1, wherein, Step S3 specifically comprises: S3-1: the deviation of the camera capturing the circle ring and the circle center is the threading and scribing error, the camera divides according to the position relationship of the measuring points in the platform coordinate system to obtain the scanning period, exposure lens and strip label to which each measuring point belongs; S3-2: subtract the platform coordinates of each measuring point actually captured by the camera from the platform coordinates of the theoretical position of each measuring point to obtain the difference between the theoretical position and the actual position of each measuring point.

4. The method for horizontal calibration error monitoring and compensation of an LDI exposure machine according to claim 1, wherein, Step S4 specifically comprises: S4-1: use the affine transformation model to fit the difference between the theoretical and actual positions of each measuring point by the least square method, and compensate the fitting coefficients to the affine transformation model respectively; S4-2: obtain the first residual value after removing the expansion, rotation error of the platform at the exposure position, and the position calibration error of the reference lens and the reference camera.

5. The method for horizontal calibration error monitoring and compensation of an LDI exposure machine according to claim 1, wherein, Step S5 specifically comprises: S5-1: according to the position relationship of the measuring points in the exposure lens coordinate system, fit the first residual value according to the different scanning periods, exposure lenses and strip labels by using a first-order linear model to obtain the lens X magnification calibration error, lens angle calibration error and lens position relationship calibration error of each exposure lens by the least square method; S5-2: compensate the lens magnification calibration error, lens angle calibration error and lens position relationship calibration error to the lens magnification and angle calibration parameters and the position relationship between the lenses respectively; S5-3: obtain the second residual value after removing the lens magnification, angle calibration of the exposure lens and the calibration error of the position relationship of the exposure lens.

6. The method for horizontal calibration error monitoring and compensation of an LDI exposure machine according to claim 1, wherein, Step S6 specifically comprises: S6-1: according to the actual platform coordinates of each measuring point, calculate the X and Y coordinates of the platform when the exposure measuring point is calculated, correspond the second residual value after removing the lens magnification, angle calibration of the exposure lens and the calibration error of the position relationship of the exposure lens to the platform position one by one, and average the grid errors of the platform position measuring points to represent the effective value of the grid error of the platform when the exposure measuring point is calculated; S6-2: The valid value of the platform position grid data is compensated in the platform grid error of the platform at the exposure position, and the remaining positions in the grid are filled using linear interpolation.

7. The method for horizontal calibration error monitoring and compensation of an LDI exposure machine according to claim 1, wherein, Also included is S7: After compensating for the error amount in steps S4-S6, re-exposure is performed after plate replacement to verify whether the compensation result is effective, and the convergence is judged according to the compensation result. S8: Calculate and compensate the calibration residual error of the alignment system.

8. The method for horizontal calibration error monitoring and compensation of an LDI exposure machine according to claim 7, wherein, In step S7, verifying whether the compensation result is effective includes: platform expansion and contraction error, reference lens and reference camera position relationship error, exposure lens magnification and angle error, exposure lens position relationship, and platform grid error at the exposure position.

9. The method for horizontal calibration error monitoring and compensation of an LDI exposure machine according to claim 7, wherein, Step S8 specifically includes: S8-1: First, expose the circular ring pattern, then select four points in the circular ring layer as alignment points to complete alignment, then expose the circle, and finally measure the X / Y center deviation of the circular ring and the circle using the alignment camera. The circular ring and the circle center error are caused by the calibration error of the alignment system; S8-2: Fit each point on the material plate to obtain the calibration residual error of the alignment system. The calibration residual error compensation parameter needs to be added when using the alignment function for subsequent exposure to reduce the overlay deviation between the circular ring and the circle center, thereby optimizing the single machine overlay performance of the machine.

Images