A method for calibrating and compensating pattern position error in laser direct writing lithography system

Abstract

The present invention relates to a method for calibrating and compensating pattern position errors in a laser direct-write lithography system, comprising the following steps: pattern position error calibration; decoupling pattern position errors using a grid calibration plate; pattern position error separation; and pattern position error compensation. Based on compensating for the low-frequency errors generated by two-axis guide rails, the present invention establishes a medium- and low-frequency position error estimation model. This method directly performs point-to-point compensation for the medium- and low-frequency errors inherent in the lithography system within the designed pattern, achieving high-precision compensation for all compensable error components. The present invention does not involve complex design and operational processes, does not rely on absolute detection technology, and improves error compensation accuracy while reducing processing costs.

Description

Technical Field

[0001] The present invention relates to the field of optical processing and detection technology, and in particular to a method for calibrating and compensating pattern position errors in a laser direct writing lithography system. Background Art

[0002] Laser direct write lithography is a mask-free micro-nanofabrication technology primarily used for the fabrication of diffraction optical elements such as computer-generated holograms, masks, microlenses, zone plates, and blazed gratings, with a manufacturing accuracy of submicron levels. The basic principle of this technology is to use a computer-controlled scanning laser beam with variable intensity to perform variable-dose exposure on the resist layer on the surface of the substrate, directly writing any designed pattern, and then forming a relief pattern on the substrate surface through a development process. The laser direct write lithography process is already mature, and companies such as Heidelberg in Germany, McConnell in Sweden, and Applied Materials in the United States have all launched commercial laser direct write lithography systems, which are widely used in space optics, optical communications, optical displays, and other fields.

[0003] Pattern position error is the deviation between the actual pattern position in the lithography process and the ideal pattern position. For diffractive optical elements, this deviation will produce wavefront error, reducing the accuracy of the diffraction wavefront. Pattern position error is mainly caused by the platform motion error of the laser direct write lithography system. Taking the Cartesian coordinate writing system as an example, the platform on which the substrate is placed is driven by a two-axis guide rail. The guide rail positioning error and orthogonality error will cause pattern position deviation, directly affecting the pattern exposure accuracy.

[0004] The golden plate calibration method is a commonly used method for calibrating pattern position errors. This method uses a high-precision grid calibration plate and executes an empty exposure calibration program to record the actual exposure position of the lithography system. The deviation between this and the ideal position coordinate is the position error generated by the platform guide rail at this location. The calibration program generates a corresponding position error matrix based on the grid calibration points. The positioning error and orthogonality error of the two-axis guide rails can be obtained through simple calculations. This calibration method has been widely used to calibrate the platform motion accuracy of commercial laser direct write lithography systems. The guide rail error calibration value can be directly compensated as a system error by adjusting the geometric relationship between the two-axis guide rails.

[0005] The gold plate calibration method of the prior art has the following disadvantages:

[0006] 1. Inability to isolate pattern errors introduced by the gold plate itself. The gold plate is processed by equipment with higher precision than the system being calibrated, yet this equipment's processing accuracy is still limited. This causes the gold plate pattern to include pattern position errors introduced by this equipment. The calibration error obtained using existing gold plate calibration techniques couples the actual pattern error with the pattern error introduced by the processing equipment, resulting in reduced calibration accuracy.

[0007] 2. Limited low-frequency error compensation. Existing gold plate calibration technology only accounts for the positioning and orthogonality errors of the two-axis guide rails. For diffractive optical elements, it can only compensate for a portion of the low-frequency wavefront errors generated in the diffracted wavefront within the calibration area. It cannot provide high-precision point-to-point compensation for errors in the entire low- and mid-frequency bands, or even the entire frequency band.

[0008] 3. The complex design process for composite phases results in limited calibration accuracy and low compensation efficiency. The composite phase CGH calibration method requires the design of a phase pattern capable of simultaneously emitting multiple wavefronts, involving composite phase superposition. This makes the simulation design process extremely complex. Furthermore, due to the complexity of the phase pattern, this verification element places high precision demands on both processing and detection technologies, relying on absolute detection technology. Summary of the Invention

[0009] The present invention aims to solve the technical problem in the prior art of laser direct writing lithography technology that the processing pattern accuracy is limited due to error coupling and incomplete compensation of low-frequency errors in the existing gold plate calibration method, and provides a calibration and compensation method for the pattern position error of the laser direct writing lithography system.

[0010] In order to solve the above technical problems, the technical solutions of the present invention are as follows:

[0011] A method for calibrating and compensating pattern position error in a laser direct writing lithography system comprises the following steps:

[0012] Step 1: Pattern position error calibration;

[0013] Use grid calibration plate to calibrate the pattern position error of laser direct writing lithography system;

[0014] Step 2: Decoupling of grid calibration pattern position error;

[0015] According to the calibration error matrix data, a grid calibration plate pattern error rotation decoupling model is established to solve the grid calibration plate pattern position error;

[0016] Step 3: Pattern position error separation;

[0017] A low-frequency position error fitting model based on affine transformation and a medium- and low-frequency position error fitting model based on optimized polynomials are respectively established to completely separate the pattern position error obtained in step 2 generated by the laser direct writing lithography system;

[0018] Step 4: Pattern position error compensation;

[0019] The position errors of each pattern separated in step 3 are compensated respectively.

[0020] In the above technical solution, step 1 is specifically as follows:

[0021] Step 1: Place the grid calibration plate on a baseplate platform driven by two-axis guide rails, and align its grid coordinate system with the system coordinate system to ensure that the subsequent calibration procedure can be executed normally;

[0022] Step 2: Execute the calibration program. The substrate platform reaches each grid point in turn according to the program design path. The system records the actual position coordinates (x a ,y a ), which is consistent with the ideal grid point coordinates (x b ,y b ) is the pattern position error at that point (x a -x b ,y a -y b ), thus obtaining the calibration error matrix P;

[0023] Step 3: Rotate the grid calibration plate 90 degrees in the plane of the substrate platform, repeat Step 1 and Step 2, and obtain the calibration error matrix Q after rotation.

[0024] In the above technical solution, step 2 is specifically as follows:

[0025] Step 1: Establish a pattern error rotation decoupling model; assume that the initial calibration error matrix is ​​P, the calibration error matrix after the grid calibration plate is rotated by a certain angle is Q, and the rotation matrix is ​​M R , the pattern position error matrix of the grid calibration plate is P S , the pattern position error matrix of the laser direct writing lithography system is P M , then there is the following relationship:

[0026] Q=M R ·P S +P M ,

[0027] P=P S +P M

[0028] Step 2: Calculate the pattern position error. According to the matrix relationship described in Step 1, the pattern position error matrices are calculated as follows:

[0029] P S =(QP) / (M R -1),

[0030] P M =PP S

[0031] Separate the pattern position error P introduced by the grid calibration plate in the calibration error matrix S, the pattern position error P generated only by the laser direct writing lithography system is obtained M .

[0032] In the above technical solution, step 3 is specifically as follows:

[0033] Step 1: Establish a low-frequency position error fitting model based on affine transformation; assume that XOY is the ideal dual-axis guide coordinate system, and X'OY' is the actual dual-axis guide coordinate system containing pattern position error; a and b are the displacements of the X'OY' coordinate system relative to the XOY coordinate system in the X-axis and Y-axis guide directions of the substrate platform, respectively; m and n are the positioning error coefficients of the X-axis and Y-axis guides, respectively; α and β are the deflection angles of the X' and Y' axes of the actual dual-axis guide relative to the X and Y axes of the ideal dual-axis guide in the plane of the substrate platform, respectively; then, the relationship between the XOY coordinate system and the X'OY' coordinate system is:

[0034]

[0035] Where x' and y' are the X'-axis and Y'-axis coordinates in the X'OY' coordinate system, and x and y are the X-axis and Y-axis coordinates in the XOY coordinate system.

[0036] Use the least squares method to find an approximate solution to the above matrix equation and solve the matrix like this:

[0037] M e =X·M

[0038] The solution is,

[0039] X=(M e ·M')·inv(M·M')

[0040] Among them, M e is the actual two-axis guide rail coordinate matrix containing the pattern position error, M is the ideal two-axis guide rail coordinate matrix, X is the matrix solution, M' is the transposed matrix of M, and inv() is the matrix inversion function;

[0041] The positioning error, orthogonal error and additional displacement generated during the calibration process of the dual-axis guide rail are solved; these low-frequency errors are extracted from the pattern position error matrix P M Separate and get the residual matrix P MR :

[0042] P MR =P M -X·M

[0043] Step 2: Establish a low-frequency position error fitting model based on optimized polynomials; assuming h i (x,y) is the optimization polynomial, A iis the polynomial coefficient, n is the number of polynomial terms, then the pattern position error matrix P MR Expressed as:

[0044]

[0045] Use the least squares method to solve the above matrix. The specific solution method is the same as in Step 1, and the polynomial coefficient matrix A is obtained. i ; The inherent low-frequency position errors of the lithography system are obtained by solving the problem, and these low-frequency position errors are extracted from the pattern position error matrix P MR Separate the pattern position residual P, which is mainly composed of high-frequency random errors. R :

[0046]

[0047] In the above technical solution, step 4 is specifically as follows:

[0048] Step 1: Two-axis guide rail error compensation: Use the guide rail error compensation program of the laser direct writing lithography system to directly compensate for the positioning error and orthogonality error of the two-axis guide rails obtained by solving;

[0049] Step 2: Systematic pattern position error compensation: Establish a pattern position error estimation model based on the calibration error data set, calculate the position error matrix that may be generated by the pattern to be processed within the calibration area; Based on the estimated pattern position error matrix, compensate for it in the coordinate point design of the pattern to be processed;

[0050] Step 3: Compensate pattern lithography processing; place the substrate to be processed in the calibration area on the two-axis guide platform and align it with the physical position of the calibration error data set; execute the laser direct writing exposure program to complete the position error compensation processing of the lithography pattern.

[0051] In the above technical solution, in Step 4, Step 2:

[0052] If the pattern position error estimation model is based on the calibration error dataset P MR , then the matrix A i is the model estimation coefficient, the estimated pattern position error P ME for:

[0053]

[0054] Among them, x e and y e They are respectively the X-axis and Y-axis coordinates of the pattern to be processed in the XOY coordinate system;

[0055] Then, the coordinate point set P of the pattern to be processed after pattern position error pre-compensation is UM Expressed as:

[0056] P UM =P I -P ME

[0057] Among them, P I is the initial coordinate point set of the pattern to be processed.

[0058] The present invention has the following beneficial effects:

[0059] The present invention's method for calibrating and compensating pattern position errors in a laser direct-write lithography system can decouple calibration grid plate pattern errors. Existing gold plate calibration technology is unable to decouple gold plate pattern errors from system pattern errors, which results in the calibration error always including the gold plate pattern error, directly affecting the accuracy of subsequent error compensation. The pattern position error calibration method proposed in the present invention establishes a pattern error rotation decoupling model by rotating the grid calibration plate, achieving high-precision decoupling between the grid calibration plate pattern error and the lithography system pattern error. On this basis, the pattern accuracy requirements for the grid calibration plate can even be appropriately reduced.

[0060] The method for calibrating and compensating the pattern position error of the laser direct writing lithography system of the present invention can achieve high-precision separation of errors in the entire frequency band. The existing gold plate calibration technology only performs fitting separation on some low-frequency components in the pattern position error, such as the positioning error and orthogonal error of the two-axis guide rails, while the remaining error still contains a large number of medium and low-frequency components. The present invention separates the low-frequency, medium-frequency and high-frequency errors of the pattern position error in turn by establishing a position error fitting model. On the basis of separating the low-frequency error generated by the two-axis guide rails, the medium and low-frequency errors and high-frequency random errors inherent in the lithography system are further fitted and separated, thereby achieving high-precision separation of errors in the entire frequency band.

[0061] The present invention's method for calibrating and compensating for pattern position errors in a laser direct-write lithography system can achieve high-precision compensation for medium- and low-frequency errors. Existing gold plate calibration technology only compensates for the positioning and orthogonal errors of two-axis guide rails, resulting in limited compensation. The present invention, based on compensating for the low-frequency errors generated by the two-axis guide rails, establishes a medium- and low-frequency position error estimation model. This model directly performs point-to-point compensation for the medium- and low-frequency errors inherent in the lithography system within the designed pattern, achieving high-precision compensation for all compensable error components.

[0062] The method for calibrating and compensating pattern position errors of a laser direct writing lithography system of the present invention does not involve complicated design processes and operation processes, does not rely on absolute detection technology, improves error compensation accuracy, and reduces processing costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0064] Figure 1 The figure is a flow chart of the method for calibrating and compensating pattern position error of the laser direct writing lithography system of the present invention. DETAILED DESCRIPTION

[0065] The inventive concept of the present invention is:

[0066] The method for calibrating and compensating pattern position errors of a laser direct writing lithography system of the present invention decouples the pattern errors introduced by the calibration grid plate, achieves high-precision compensation of pattern position errors in the calibration area, and greatly improves the pattern position accuracy of the laser direct writing lithography system.

[0067] In the pattern position error calibration stage, the present invention establishes a grid calibration plate pattern error rotation decoupling model. By calibrating the pattern position error of the grid plate at different rotation angles, high-precision decoupling between the pattern position error of the grid calibration plate and the pattern position error of the laser direct writing lithography system is achieved.

[0068] During the pattern position error separation phase, the present invention first establishes a low-frequency position error fitting model based on affine transformation to separate the low-frequency pattern position errors caused by the positioning errors and orthogonal errors of the two-axis guideways. Secondly, a medium- and low-frequency position error fitting model based on an optimized polynomial is established. By analyzing the distribution trend of the pattern position residuals, the medium- and low-frequency errors inherent in the laser direct write lithography system are further separated. Ultimately, a high-frequency pattern position residual consisting solely of random errors is obtained.

[0069] During the pattern position error compensation phase, the present invention establishes a pattern position error estimation model based on a calibration error dataset. This model directly calculates the pattern position error introduced by the laser direct write lithography system based on the actual pattern to be processed, and then feeds this into the design phase of the pattern to be processed. By pre-compensating the pattern position error of the processed pattern, the system achieves high-precision, full-band, and high-efficiency compensation of the laser direct write lithography system's position error.

[0070] The present invention will be described in detail below with reference to the accompanying drawings.

[0071] Focusing on the calibration, analysis and compensation of pattern position errors in a laser direct writing lithography system, the present invention provides a method for calibrating and compensating pattern position errors in a laser direct writing lithography system, including the following steps:

[0072] 1. Pattern position error calibration. Use a grid calibration plate to calibrate the pattern position error of the laser direct write lithography system. The basic calibration process is as follows:

[0073] Step 1: Place the grid calibration plate on a baseplate platform driven by two-axis guide rails, and align its grid coordinate system with the system coordinate system to ensure that the subsequent calibration procedure can be executed normally;

[0074] Step 2: Execute the calibration program. The substrate platform reaches each grid point in turn according to the program design path. The system records the actual position coordinates (x a ,y a ), which is consistent with the ideal grid point coordinates (x b ,y b ) is the pattern position error at that point (x a -x b ,y a -y b ), thus obtaining the calibration error matrix P;

[0075] Step 3: Rotate the grid calibration plate 90 degrees in the plane of the substrate platform, repeat Step 1 and Step 2, and obtain the calibration error matrix Q after rotation.

[0076] 2. Decoupling of grid calibration plate pattern position error. Based on the calibration error matrix data, a grid calibration plate pattern error rotation decoupling model is established to solve the grid plate pattern position error. The basic solution steps are as follows:

[0077] Step 1: Establish a pattern error rotation decoupling model. Assume that the initial calibration error matrix is ​​P, the calibration error matrix after the grid calibration plate is rotated by a certain angle is Q, and the rotation matrix is ​​M R , the pattern position error matrix of the grid calibration plate is P S , the pattern position error matrix of the laser direct writing lithography system is P M , then there is the following relationship:

[0078]

[0079] Step 2: Calculate the pattern position error. According to the matrix relationship described in Step 1, the pattern position error matrices are calculated as follows:

[0080]

[0081] Separate the pattern position error P introduced by the grid calibration plate in the calibration error matrix S , the pattern position error P generated only by the laser direct writing lithography system is obtained M .

[0082] 3. Pattern position error separation. A low-frequency position error fitting model based on affine transformation and a medium- and low-frequency position error fitting model based on optimized polynomials are established to completely separate the pattern position error generated by the laser direct write lithography system. The basic separation steps are as follows:

[0083] Step 1: Establish a low-frequency position error fitting model based on affine transformation. This model is mainly used to separate the low-frequency errors introduced by the positioning error and orthogonal error of the two-axis guide rails. Assume that XOY is the ideal two-axis guide rail coordinate system, and X'OY' is the actual two-axis guide rail coordinate system containing the pattern position error. a and b are the displacements of the X'OY' coordinate system relative to the XOY coordinate system in the X-axis and Y-axis guide directions of the substrate platform, respectively. m and n are the positioning error coefficients of the X-axis and Y-axis guide rails, respectively. α and β are the deflection angles of the X'axis and Y'axis of the actual two-axis guide rail relative to the X-axis and Y-axis of the ideal two-axis guide rail in the plane of the substrate platform. Then, the relationship between the XOY coordinate system and the X'OY' coordinate system can be described as

[0084]

[0085] Where x' and y' are the X'-axis and Y'-axis coordinates in the X'OY' coordinate system, and x and y are the X-axis and Y-axis coordinates in the XOY coordinate system.

[0086] Use the least squares method to find an approximate solution to the above matrix equation, that is, to solve the following matrix:

[0087] M e =X·M (4)

[0088] The solution is,

[0089] X=(M e ·M')·inv(M·M') (5)

[0090] Among them, M e is the actual dual-axis guide rail coordinate matrix containing the pattern position error, M is the ideal dual-axis guide rail coordinate matrix, X is the matrix solution, M' is the transposed matrix of M, and inv() is the matrix inversion function.

[0091] This can solve the positioning error (i.e. m, n), orthogonal error (i.e. α-β) of the dual-axis guide rail, and the additional displacement (i.e. a, b) generated during the calibration process. These low-frequency errors are extracted from the pattern position error matrix P M Separate and get the residual matrix P MR :

[0092] P MR =P M -X·M (6)

[0093] Step 2: Establish a low-frequency and medium-frequency position error fitting model based on the optimization polynomial. This model is mainly used to separate the low-frequency and medium-frequency errors inherent in the lithography system. Assume h i (x,y) is the optimization polynomial, A i is the polynomial coefficient, n is the number of polynomial terms, then the pattern position error matrix PMR It can be expressed as:

[0094]

[0095] Use the least squares method to solve the above matrix. The specific solution method is the same as in Step 1, and the polynomial coefficient matrix A is obtained. i The inherent low- and medium-frequency position errors of the lithography system can be solved and these low- and medium-frequency position errors can be extracted from the pattern position error matrix P MR Separate the pattern position residual P, which is mainly composed of high-frequency random errors. R :

[0096]

[0097] 4. Pattern position error compensation. Compensate the position errors of each separated pattern separately. The basic compensation steps are as follows:

[0098] Step 1: Two-axis guide rail error compensation. Use the guide rail error compensation program of the laser direct writing lithography system to directly compensate for the positioning error and orthogonality error of the two-axis guide rails obtained by calculation;

[0099] Step 2: Systematic pattern position error compensation. Build a pattern position error estimation model based on the calibration error dataset, calculate the position error matrix that may be generated by the pattern to be processed within the calibration area, and compensate for the estimated pattern position error matrix in the coordinate design of the pattern to be processed.

[0100] Step 3: Compensate Pattern Lithography. Place the substrate to be processed within the calibration area on the two-axis guide platform and align it with the physical position of the calibration error dataset. Execute the laser direct write exposure process to complete the position error compensation of the lithographic pattern.

[0101] The pattern position error estimation model is the key to the system pattern position error compensation. If the model is based on the calibration error data set P MR , then the matrix A i That is the model estimation coefficient, the estimated pattern position error P ME for:

[0102]

[0103] Among them, x e and y e They are respectively the X-axis and Y-axis coordinates of the pattern to be processed in the XOY coordinate system.

[0104] Then, the coordinate point set P of the pattern to be processed after pattern position error pre-compensation is UM It can be expressed as:

[0105] PUM =P I -P ME (10)

[0106] Among them, P I is the initial coordinate point set of the pattern to be processed.

[0107] Simulation, experiments and use all show that the method for calibrating and compensating the pattern position error of the laser direct writing lithography system of the present invention is practical and effective.

[0108] In the above-mentioned specific embodiments, the grid calibration plate used in the calibration method of the present invention can be any tool with a calibration function, including but not limited to a flat plate, an inclined plate, a curved plate, etc. The calibration points (or targets) used for calibration can be of any shape (including but not limited to a cross, a circle, a square, a diamond, etc.), in any form (including but not limited to grooves, protrusions, etc.), and in any distribution form (including but not limited to a grid distribution, a ring distribution, a random distribution, etc.).

[0109] In the above specific implementation, the calibration method of the present invention partially mentions rotating the grid calibration plate by 90 degrees in Step 3. This angle can be any angle, and only the subsequent error decoupling model needs to be modified accordingly.

[0110] In the above specific embodiment, the grid calibration plate pattern position error decoupling method of the present invention can also be simply replaced by the "low-frequency position error fitting model based on affine transformation" proposed in the present invention. The specific replacement method is:

[0111] Step 1: Use the low-frequency position error fitting model based on affine transformation to solve the pattern position error matrix before and after the calibration plate is rotated 90 degrees, and obtain the positioning errors of the dual-axis guide rail before and after the rotation as m0, n0 and m0 respectively. 90 、n 90 The deflection angles of the dual-axis guide rails before and after rotation are α0, β0 and α 90 , β 90 ;

[0112] Step 2: Assume that m1 and n1 are the positioning error coefficients of the X-axis and Y-axis guide rails of the lithography system, α1 and β1 are the deflection angles of the X-axis and Y-axis of the actual dual-axis guide rail of the lithography system relative to the ideal state; m2 and n2 are the positioning error coefficients of the X-axis and Y-axis guide rails introduced by the grid calibration plate, α2 and β2 are the deflection angles of the X-axis and Y-axis of the actual dual-axis guide rail introduced by the grid calibration plate relative to the ideal state, then the error matrix of the grid calibration plate at 0 degrees and 90 degrees has the following relationship:

[0113]

[0114] Solving the above matrix equations yields the pattern position error introduced by the grid calibration plate and the pattern position error generated by the lithography system. This method can only decouple the low-frequency position error introduced by the grid calibration plate.

[0115] Similarly, the "90-degree rotation of the calibration plate" mentioned in Step 1 can be any angle, and the subsequent error decoupling model only needs to be modified accordingly.

[0116] In the above specific embodiments, the basic separation steps mentioned in the pattern position error separation part of the present invention do not constitute a limitation on the separation order, that is, the separation order of low-frequency, medium-low-frequency, and high-frequency errors can be arbitrary.

[0117] In the above specific implementation manner, the least squares method mentioned in the technical implementation scheme of the present invention is only a solution method, and the solution method of replacing the matrix equation cannot be considered as a new technical solution different from the present invention.

[0118] The present invention's method for calibrating and compensating pattern position errors in a laser direct-write lithography system can decouple calibration grid plate pattern errors. Existing gold plate calibration technology is unable to decouple gold plate pattern errors from system pattern errors, which results in the calibration error always including the gold plate pattern error, directly affecting the accuracy of subsequent error compensation. The pattern position error calibration method proposed in the present invention establishes a pattern error rotation decoupling model by rotating the grid calibration plate, achieving high-precision decoupling between the grid calibration plate pattern error and the lithography system pattern error. On this basis, the pattern accuracy requirements for the grid calibration plate can even be appropriately reduced.

[0119] The method for calibrating and compensating the pattern position error of the laser direct writing lithography system of the present invention can achieve high-precision separation of errors in the entire frequency band. The existing gold plate calibration technology only performs fitting separation on some low-frequency components in the pattern position error, such as the positioning error and orthogonal error of the two-axis guide rails, while the remaining error still contains a large number of medium and low-frequency components. The present invention separates the low-frequency, medium-frequency and high-frequency errors of the pattern position error in turn by establishing a position error fitting model. On the basis of separating the low-frequency error generated by the two-axis guide rails, the medium and low-frequency errors and high-frequency random errors inherent in the lithography system are further fitted and separated, thereby achieving high-precision separation of errors in the entire frequency band.

[0120] The present invention's method for calibrating and compensating for pattern position errors in a laser direct-write lithography system can achieve high-precision compensation for medium- and low-frequency errors. Existing gold plate calibration technology only compensates for the positioning and orthogonal errors of two-axis guide rails, resulting in limited compensation. The present invention, based on compensating for the low-frequency errors generated by the two-axis guide rails, establishes a medium- and low-frequency position error estimation model. This model directly performs point-to-point compensation for the medium- and low-frequency errors inherent in the lithography system within the designed pattern, achieving high-precision compensation for all compensable error components.

[0121] The method for calibrating and compensating pattern position errors of a laser direct writing lithography system of the present invention does not involve complicated design processes and operation processes, does not rely on absolute detection technology, improves error compensation accuracy, and reduces processing costs.

[0122] Obviously, the above embodiments are merely examples for clarity of explanation and are not intended to limit the implementation methods. Those skilled in the art will readily appreciate that other variations or modifications based on the above descriptions are possible. It is not necessary and impossible to enumerate all implementation methods here. Obvious variations or modifications arising therefrom remain within the scope of protection of the present invention.

Claims

1. A method for calibrating and compensating pattern position error in a laser direct writing lithography system, characterized in that: The following steps are involved: Step 1: Pattern position error calibration; Calibrate the pattern position error of the laser direct write lithography system using a grid calibration plate, specifically including the following steps: Step 1: Place the grid calibration plate on a baseplate platform driven by two-axis guide rails, and align its grid coordinate system with the system coordinate system to ensure that the subsequent calibration procedure can be executed normally; Step 2: Execute the calibration program. The substrate platform reaches each grid point in turn according to the program design path. The system records the actual position coordinates (x a ,y a ), which is consistent with the ideal grid point coordinates (x b ,y b ) is the pattern position error at that point (x a -x b ,y a -y b ), thus obtaining the calibration error matrix P; Step 3: Rotate the grid calibration plate 90 degrees in the plane of the substrate platform, repeat Step 1 and Step 2 to obtain the calibration error matrix Q after rotation; Step 2: Decoupling of grid calibration pattern position error; According to the calibration error matrix data, a grid calibration plate pattern error rotation decoupling model is established to solve the grid calibration plate pattern position error, which specifically includes the following steps: Step 1: Establish a pattern error rotation decoupling model; according to the initial calibration error matrix P obtained in step 1, the calibration error matrix Q after rotation, the rotation matrix is ​​M R , the pattern position error matrix of the grid calibration plate is P S , the pattern position error matrix of the laser direct writing lithography system is P M , then there is the following relationship: Q=M R ·P S +P M , P=P S +P M Step 2: Calculate the pattern position error. According to the matrix relationship described in Step 1, the pattern position error matrices are calculated as follows: P S =(Q-P) / (M R -1), P M =P-P S Separate the pattern position error P introduced by the grid calibration plate in the calibration error matrix S , the pattern position error P generated only by the laser direct writing lithography system is obtained M ; Step 3: Pattern position error separation; A low-frequency position error fitting model based on affine transformation and a medium- and low-frequency position error fitting model based on optimized polynomials are respectively established to completely separate the pattern position error obtained in step 2 generated by the laser direct writing lithography system; Step 4: Pattern position error compensation; The position errors of each pattern separated in step 3 are compensated respectively.

2. The method for calibrating and compensating pattern position error of a laser direct writing lithography system according to claim 1, characterized in that: Step 3 is as follows: Step 1: Establish a low-frequency position error fitting model based on affine transformation; assume that XOY is the ideal dual-axis guide coordinate system, and X'OY' is the actual dual-axis guide coordinate system containing pattern position error; a and b are the displacements of the X'OY' coordinate system relative to the XOY coordinate system in the X-axis and Y-axis guide directions of the substrate platform, respectively; m and n are the positioning error coefficients of the X-axis and Y-axis guides, respectively; α and β are the deflection angles of the X' and Y' axes of the actual dual-axis guide relative to the X and Y axes of the ideal dual-axis guide in the plane of the substrate platform, respectively; then, the relationship between the XOY coordinate system and the X'OY' coordinate system is: Where x' and y' are the X'-axis and Y'-axis coordinates in the X'OY' coordinate system, and x and y are the X-axis and Y-axis coordinates in the XOY coordinate system. Use the least squares method to find an approximate solution to the above matrix equation and solve the matrix like this: M e =X·M The solution is, X=(M e ·M')·inv(M·M') Among them, M e is the actual two-axis guide rail coordinate matrix containing the pattern position error, M is the ideal two-axis guide rail coordinate matrix, X is the matrix solution, M' is the transposed matrix of M, and inv() is the matrix inversion function; The positioning error, orthogonal error and additional displacement generated during the calibration process of the dual-axis guide rail are solved; these low-frequency errors are extracted from the pattern position error matrix P M Separate and get the residual matrix P MR : P MR =P M -X·M Step 2: Establish a low-frequency position error fitting model based on optimized polynomials; assuming h i (x,y) is the optimization polynomial, A i is the polynomial coefficient, n is the number of polynomial terms, then the pattern position error matrix P MR Expressed as: Use the least squares method to solve the above matrix. The specific solution method is the same as in Step 1, and the polynomial coefficient matrix A is obtained. i ; The inherent low-frequency position errors of the lithography system are obtained by solving the problem, and these low-frequency position errors are extracted from the pattern position error matrix P MR Separate the pattern position residual P, which is mainly composed of high-frequency random errors. R :

3. The method for calibrating and compensating pattern position error of a laser direct writing lithography system according to claim 1, wherein: Step 4 is as follows: Step 1: Two-axis guide rail error compensation: Use the guide rail error compensation program of the laser direct writing lithography system to directly compensate for the positioning error and orthogonality error of the two-axis guide rails obtained by solving; Step 2: Systematic pattern position error compensation: Establish a pattern position error estimation model based on the calibration error data set, calculate the position error matrix that may be generated by the pattern to be processed within the calibration area; Based on the estimated pattern position error matrix, compensate for it in the coordinate point design of the pattern to be processed; Step 3: Compensate pattern lithography processing; place the substrate to be processed in the calibration area on the two-axis guide platform and align it with the physical position of the calibration error data set; execute the laser direct writing exposure program to complete the position error compensation processing of the lithography pattern.

4. The method for calibrating and compensating pattern position error of a laser direct writing lithography system according to claim 3, wherein: In Step 4, Step 2: If the pattern position error estimation model is based on the calibration error dataset P MR , then the matrix A i is the model estimation coefficient, the estimated pattern position error P ME for: Among them, x e and y e They are respectively the X-axis and Y-axis coordinates of the pattern to be processed in the XOY coordinate system; Then, the coordinate point set P of the pattern to be processed after pattern position error pre-compensation is UM Expressed as: P UM =P I -P ME Among them, P I is the initial coordinate point set of the pattern to be processed.

Images