OPC model data weight setting method

Abstract

The present invention discloses a method for setting OPC model data weights, comprising: step 1, classifying graphics on a layout and setting reference weights for each type of graphics; step 2, setting weights for each data point in the OPC model, including: calculating a first coefficient, the first coefficient being a coefficient related to the line width of the graphics at the data point; calculating a second coefficient, the second coefficient being a coefficient related to the reliability of the measurement size of the graphics at the data point; and calculating the weight of the data point, the weight of the data point being the product of the reference weight of the graphics category to which the data point belongs, the first coefficient, and the second coefficient. The present invention can effectively reduce model error and root mean square value, thereby improving OPC model accuracy.

Description

Technical Field

[0001] The present invention relates to a semiconductor integrated circuit manufacturing method, and in particular to a method for setting optical proximity correction (OPC) model data weights. Background Art

[0002] Model-based OPC correction methods play a very important role in the field of resolution enhancement technology. OPC tools can reduce edge position error (EPE) by moving mask edges. Therefore, the mask correction results depend largely on the accuracy of the OPC model. As one of the many OPC tools, the OPC model is actually semi-empirical. Apart from many physical and optical parameters, wafer data is very critical for modeling and calibration. To some extent, the OPC model establishment and calibration process is actually adjusting the model data weight, and the weight adjustment is often affected by the experience of the modeling engineer.

[0003] like Figure 1 As shown, it is a schematic diagram of the existing model-based OPC correction; first, a target layout 101 is provided. The target layout 101 is a design layout formed according to design requirements. The target layout 101 has a pattern 105a, and the wafer pattern finally formed on the wafer needs to be close to the pattern 105a.

[0004] Secondly, an OPC model 102 is required. The OPC model 102 fully describes the entire photolithography process, including the optical system, mask, photoresist, and etching process, including the optical model and the photoresist photochemical reaction model. Typically, the OPC model 102 used in OPC correction is a semi-empirical simplified photolithography model.

[0005] Next, OPC model 102 and target layout 101 are combined to perform OPC correction as shown in mark 103 to obtain mask layout 104. Pattern 105b on mask layout 104 is adjusted based on pattern 105a so that the pattern transferred to the wafer after the photolithography process is close to the target pattern 105a.

[0006] In order to improve the accuracy of OPC correction, the OPC model 102 needs to be calibrated and verified. Generally, the process of calibrating the OPC model 102 is mainly the process of adjusting the weights of each data point in the OPC model 102. Each data point of the OPC model 102 records the corresponding graphic data. During OPC correction, when the graphics need to be adjusted, they are adjusted according to the size of the weight setting, so that the adjustment size of the graphics corresponding to different data points is different, and the EPE of the graphics corresponding to each data point is reduced. In existing methods, weight adjustment is often affected by the experience of the modeling engineer, which will affect the accuracy of the OPC model. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to provide a method for setting the weight of OPC model data, which can effectively reduce the model error and the root mean square value and improve the accuracy of the OPC model.

[0008] To solve the above technical problems, the present invention provides an OPC model data weight setting method comprising the following steps:

[0009] Step 1: Classify the graphics on the layout and set reference weights for each type of graphics;

[0010] Step 2: Set the weight of each data point in the OPC model, including:

[0011] Calculating a first coefficient, where the first coefficient is a coefficient related to a graphic line width size of the data point;

[0012] calculating a second coefficient, the second coefficient being a coefficient related to the reliability of the measurement size of the pattern of the data points;

[0013] The weight of the data point is calculated, where the weight of the data point is a product of the reference weight of the graphic category to which the graphic of the data point belongs, the first coefficient of the data point, and the second coefficient of the data point.

[0014] A further improvement is that, in step 1, the category of the graphic includes an anchor point, and the line width size of the anchor point is a first line width size.

[0015] A further improvement is that, in step 2, the smaller the difference between the graphic line width of the data point and the first line width, the larger the first coefficient.

[0016] A further improvement is that the formula for calculating the first coefficient is:

[0017]

[0018] Wherein, Kcd represents the first coefficient, CD αDenotes the first line width dimension, CD i represents the graphic line width of the data point, and i represents the number of the data point in the data set of the OPC model.

[0019] CD a and CD i All dimensions are measured from the wafer.

[0020] A further improvement is that the characterization value related to the reliability of the measurement size of the graph of the data points is 3*sigma, where sigma represents the standard deviation.

[0021] A further improvement is that the formula for calculating the second coefficient is:

[0022]

[0023] Here, Kq represents the second coefficient.

[0024] A further improvement is that the measured size of the pattern of the data points is obtained by measuring the photoresist pattern transferred onto the wafer.

[0025] A further improvement is that, in step 1, the categories of the graphics further include 1D graphic structure (1D pattern), 1.5D graphic structure (1.5D pattern) and 2D graphic structure (2D pattern).

[0026] A further improvement is that the data set of the OPC model includes more than several thousand data points.

[0027] Based on the reference weights of various types of graphics formed according to graphic classification, the present invention also modifies the reference weights according to the measured size of the graphics of each data point, the deviation from the anchor point, and the reliability of the measurement value to form the final weight. During the OPC model establishment and calibration process, since the weight reflects the reliability and importance of the measured size of each data point, it can effectively reduce the model error and the root mean square value, thereby improving the accuracy of the OPC model. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0029] Figure 1 It is a schematic diagram of the existing model-based OPC correction;

[0030] Figure 2 It is a flow chart of a method for setting OPC model data weights according to an embodiment of the present invention. DETAILED DESCRIPTION

[0031] like Figure 2 FIG. 1 is a flow chart of a method for setting weights of OPC model data according to an embodiment of the present invention. The method for setting weights of OPC model data according to an embodiment of the present invention comprises the following steps:

[0032] Step 1: Classify the graphics on the layout and set reference weights for each type of graphics.

[0033] In the embodiment of the present invention, the category of the graphic includes an anchor point, and the line width size of the anchor point is a first line width size.

[0034] The graphics categories also include 1D graphics structure, 1.5D graphics structure and 2D graphics structure.

[0035] The OPC model includes a data set corresponding to the graphics of the layout, wherein the data set includes a plurality of data points, each of which records parameters of the corresponding graphics. The number of data points included in the data set is more than several thousand, such as more than 10,000.

[0036] Step 2: Set the weight of each data point in the OPC model, including:

[0037] Calculating a first coefficient, where the first coefficient is a coefficient related to a graphic line width size of the data point;

[0038] calculating a second coefficient, the second coefficient being a coefficient related to the reliability of the measurement size of the pattern of the data points;

[0039] The weight of the data point is calculated, where the weight of the data point is a product of the reference weight of the graphic category to which the graphic of the data point belongs, the first coefficient of the data point, and the second coefficient of the data point.

[0040] In the embodiment of the present invention, the smaller the difference between the graphic line width of the data point and the first line width, the larger the first coefficient.

[0041] In a preferred embodiment, the formula for calculating the first coefficient is:

[0042]

[0043] Wherein, Kcd represents the first coefficient, CD α Denotes the first line width dimension, CD i represents the graphic line width of the data point, and i represents the number of the data point in the data set of the OPC model, that is, i in the i-th data point.

[0044] CD a and CD i All dimensions are measured from the wafer.

[0045] In the embodiment of the present invention, a characterization value related to the reliability of the measurement size of the graph of the data points is 3*sigma, where sigma represents the standard deviation.

[0046] The measured size of the pattern of the data points is obtained by measuring the photoresist pattern transferred onto the wafer.

[0047] In a preferred embodiment, the formula for calculating the second coefficient is:

[0048]

[0049] Here, Kq represents the second coefficient.

[0050] Combining formulas (1) and (2), we can see that the formula for calculating the weight of the data point is:

[0051]

[0052] Among them, w i Represents the weight of the data point, w ref The reference weight represents the graphic category to which the graphic of the data point belongs.

[0053] In some embodiments, the anchor's w ref Set to 100, 1D pattern w ref Set to 10, 1.5D pattern w ref Set to 3, 2D pattern w ref Set to 1.

[0054] In the embodiment of the present invention, based on the reference weight formed according to the graphic classification, the reference weight is also corrected according to the measured size of the graphic of each data point to form the final weight. In the OPC correction, since the weight reflects the measured size of each data point, it can eliminate the adverse effects brought by the experience of the modeling engineer, thereby effectively reducing the model error and the root mean square value, thereby improving the accuracy of the OPC model.

[0055] The following describes the application of the OPC model data weight setting method according to an embodiment of the present invention to the weight setting of the layout pattern model of the first metal layer, namely, the M1BB model. The weight setting of the M1BB model includes:

[0056] Collect and organize all measurement data;

[0057] Set the anchor, 1D, 1.5D and 2D pattern w respectively ref They are 100, 10, 3 and 1 respectively;

[0058] According to formula (3), the new weight of each data point is calculated as w i ;

[0059] Using the new weight data to calibrate the model is to calibrate the M1BB model.

[0060] The method of the embodiment of the present invention can ultimately obtain a high-precision model whose model error basically meets the requirements. As shown in Table 1, the present invention can effectively reduce the root mean square (RMS) and mean (mean) value of the model error, and the inspec ratio of the advanced process model with a total data point of more than 10,000 can be increased to more than 95%.

[0061] Table 1

[0062]

[0063] The present invention has been described in detail above by means of specific embodiments, but these do not constitute limitations of the present invention. Without departing from the principles of the present invention, those skilled in the art may make many variations and improvements, which should also be considered as the scope of protection of the present invention.

Claims

1. A method for setting weights of OPC model data, characterized in that: The steps include: Step 1: Classify the graphics on the layout and set reference weights for each type of graphics; The category of the graphic includes an anchor point, and the line width size of the anchor point is a first line width size; Step 2: Set the weight of each data point in the OPC model, including: Calculating a first coefficient, where the first coefficient is a coefficient related to a graphic line width size of the data point; The smaller the difference between the graphic line width of the data point and the first line width, the larger the first coefficient; The formula for calculating the first coefficient is: Wherein, Kcd represents the first coefficient, CD α Denotes the first line width dimension, CD i represents the line width of the graphic of the data point, i represents the number of the data point in the data set of the OPC model; calculating a second coefficient, the second coefficient being a coefficient related to the reliability of the measurement size of the graphic of the data point; A characterization value related to the reliability of the measured size of the data point pattern is 3*sigma, the measured size of the data point pattern is obtained by measuring a photoresist pattern transferred onto a wafer, the photoresist pattern transferred onto the wafer includes a plurality of data point patterns, and sigma represents a standard deviation of the measured size of the data point pattern obtained by measuring the plurality of photoresist patterns; The formula for calculating the second coefficient is: Wherein, Kq represents the second coefficient; The weight of the data point is calculated, where the weight of the data point is a product of the reference weight of the graphic category to which the graphic of the data point belongs, the first coefficient of the data point, and the second coefficient of the data point.

2. The method for setting OPC model data weights according to claim 1, wherein: CD a and CD i All dimensions are measured from the wafer.

3. The OPC model data weight setting method according to claim 1, wherein: In step 1, the graphic categories further include 1D graphic structure, 1.5D graphic structure and 2D graphic structure.

4. The method for setting OPC model data weights according to claim 1, wherein: The data set of the OPC model includes more than several thousand data points.

Images