Imaging apparatus and imaging method
The imaging device corrects for thermal expansion of both the table and sample using temperature sensors and laser interferometers, ensuring precise positioning and high-throughput imaging of semiconductor circuits.
Patent Information
- Application Number
- JP2023080687
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-05-16
- Publication Date
- 2024-11-28
AI Technical Summary
Existing imaging technologies for highly integrated semiconductor circuits fail to accurately image samples due to thermal expansion of the sample itself, which is not accounted for in previous correction methods, leading to poor positional accuracy and reduced throughput.
An imaging device and method that includes a temperature sensor, chuck, laser interferometers, and a control unit to calculate and correct for thermal expansion of both the table and sample, ensuring precise positioning by measuring and adjusting the position of the electron beam based on temperature changes.
The solution enables accurate imaging of samples regardless of thermal expansion, improving positional accuracy and allowing high-throughput inspection by reducing deviations between the intended and actual photographed areas.
Smart Images

Figure 2024164941000001_ABST
Abstract
Description
[Technical field]
[0001] The present invention relates to an imaging device and an imaging method. [Background technology]
[0002] In recent years, the integration density of semiconductor products has been increasing, and there is a demand for even higher resolution circuit patterns. For samples on which circuit patterns are formed, such as semiconductor wafers, various inspection methods are used for the purpose of quality control and yield improvement. For example, a technique is known in which a semiconductor wafer is photographed using a scanning electron microscope, and the circuit pattern is inspected from the obtained image.
[0003] In order to achieve high integration of semiconductors, the line width of the circuit pattern is being reduced, and in order to photograph and inspect it, it is necessary to photograph it at a high magnification, which requires narrowing the field of view. To narrow the field of view to be photographed, a field of view position accuracy that matches it is required. In other words, if the field of view position accuracy is poor and the position error is larger than the field of view range, the photographed location will be outside the intended field of view range. Therefore, in a photographing device for inspecting highly integrated circuits, it is important to increase the field of view position accuracy.
[0004] In a scanning electron microscope, an image is taken by scanning an electron beam, so for a sample whose position is fixed, it is possible to change the image taking location by changing the scanning position of the electron beam. Therefore, it is possible to improve the positional accuracy of the image taking location by taking an image once, analyzing the image, investigating the deviation between the target position and the actual image taking location, and then taking an image a second time after correcting the amount of positional deviation.
[0005] On the other hand, in the imaging device used in the manufacturing process of semiconductor products, it is also required to increase the throughput in order to improve productivity. For this reason, it is expected that the time required to take an image of one location can be reduced by taking an image only once, rather than taking an image twice, and that this will increase the throughput. Even if two images are taken, if the field of view position accuracy in the first image is too poor, it will be difficult to determine which part of the highly integrated circuit pattern has been photographed, and it will be impossible to correct the position and take the second image. For this reason, in order to prevent losing track of the field of view position of the imaged location, it is important to improve the position accuracy in the stage before taking the first image.
[0006] There are various physical phenomena that cause position errors in the imaging area, and they are all intertwined in a complex way. In particular, the stage that positions the sample generates heat due to the motor in the moving mechanism and frictional heat, and the heat spreads and dissipates due to thermal conduction and radiation of solids. In addition, the contact points of the stage change as it moves, and the heat transfer state is constantly changing due to changes in the relative distance to the surrounding objects, which changes the temperature distribution and causes thermal deformation.
[0007] One approach to unraveling such complex phenomena and improving the positional accuracy of the shooting location is, for example, Patent Document 1. Patent Document 1 discloses a technology for controlling the stage position by taking into consideration that the mirror shape changes depending on the temperature in a stage device that uses a mirror and a laser interferometer to measure the position, and correcting the resulting position measurement error (paragraph 0023, etc.).
[0008] Additionally, Patent Document 2 states in paragraph 0151 that "the amount of thermal deformation (variation in linear expansion) of substrate P caused by the temperature difference between the alignment process and the immersion exposure process may be calculated in advance, and a correction amount for correcting that variation may be calculated, and based on that correction amount, the positional relationship between substrate P and the pattern image when performing overlay exposure may be corrected." [Prior art documents] [Patent documents]
[0009] [Patent Document 1] JP 2002-118050 A [Patent Document 2] JP 2017-129875 A Summary of the Invention [Problem to be solved by the invention]
[0010] However, the techniques of Patent Documents 1 and 2 only take into account the thermal expansion of the stage device and substrate on which the sample is mounted when correcting the position of the imaging location, so there is a problem in that if there is thermal expansion in the sample itself, it is not possible to perform imaging with high accuracy.
[0011] SUMMARY OF THE PRESENT EMBODIMENTS Accordingly, an object of the present invention is to provide an imaging device and an imaging method that are capable of accurately imaging a sample regardless of the presence or absence of thermal expansion of the sample. [Means for solving the problem]
[0012] In order to solve the above problem, the imaging device of the present invention includes, for example, a table, a chuck whose back surface is fixed to the table and which adsorbs a sample on its front surface, a temperature sensor that acquires the temperature of the chuck, a column that images the sample mounted on the chuck, a mirror provided on the table, a laser interferometer that irradiates the mirror with laser light and measures the distance to the mirror, and a control unit that calculates a position correction amount of the imaging location based on the temperature of the chuck, wherein the control unit calculates the temperature of the table based on the temperature of the chuck, calculates the amount of thermal expansion of the table based on the temperature of the table, calculates the amount of thermal expansion of the sample based on the temperature of the chuck, and calculates the amount of position correction of the imaging location based on the amount of thermal expansion of the table and the amount of thermal expansion of the sample.
[0013] Furthermore, the imaging method of the present invention is, for example, an imaging method for imaging a sample mounted on a chuck, and includes a table temperature calculation step of calculating the temperature of a table to which the chuck is fixed based on the temperature of the chuck, a table thermal expansion amount calculation step of calculating the thermal expansion amount of the table based on the temperature of the table, a sample thermal expansion amount calculation step of calculating the thermal expansion amount of the sample based on the temperature of the chuck, a position correction amount calculation step of calculating a position correction amount of the imaging location based on the thermal expansion amount of the table and the thermal expansion amount of the sample, and a position correction step of correcting the position of the imaging location based on the position correction amount. Effect of the Invention
[0014] According to the present invention, it is possible to provide an imaging device and an imaging method that can accurately image a sample regardless of the presence or absence of thermal expansion of the sample. [Brief description of the drawings]
[0015] [Figure 1] Basic configuration of the imaging device [Diagram 2] FIG. 1 is a diagram for explaining a coordinate system in a photographing device. [Diagram 3] Operation flow diagram of the photographing device in the first embodiment [Figure 4] Detailed flow diagram of step S4 in Example 1 [Diagram 5] Detailed flow diagram of step S4 in the second embodiment [Figure 6] Operation flow diagram of the photographing device in the third embodiment [Figure 7] Detailed flow diagram of step S4 in the third embodiment [Figure 8] Detailed flow diagram of step S4 in the fourth embodiment DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0016] The present invention will be described in detail below with reference to the drawings. Note that the present invention is not limited to the embodiments described below. These embodiments are merely illustrative, and the present invention can be embodied in various modified and improved forms based on the knowledge of those skilled in the art. In addition, in each of the drawings used in the following description, the same reference numerals are used for common devices and equipment, and descriptions of devices, equipment, and operations that have already been described may be omitted. EXAMPLES
[0017] Fig. 1 shows the basic configuration of the imaging device 100 in this embodiment. Fig. 1 is a side view of the imaging device 100. Fig. 2 illustrates a coordinate system of position coordinates in order to explain the field of view position error (positional deviation of the imaging location) of the sample that is the subject of discussion in this invention. Fig. 2 is a top view of the imaging device 100. Fig. 1 corresponds to a side view seen from the positive direction of the Y axis in the coordinate system shown in Fig. 2.
[0018] <Overall configuration of the imaging device 100> As shown in FIG. 1, in the imaging device 100, a table 105 is present on the top stage of an XY stage 108 (not shown). The XY stage 108 is a stage that can move in the X-axis direction and the Y-axis direction in FIG. 2, but may also be movable in other directions. Furthermore, a chuck 103 is present on the table 105, and the back surface of this chuck 103 is fixed to the table 105 through a plurality of chuck supports 107. It is preferable to align the center of gravity of the portion where the plurality of chuck supports 107 are connected to the chuck 103 with the center of the chuck 103. For example, when there are three chuck supports 107, the connection points are arranged in an equilateral triangle, and the center of gravity of the triangle is aligned with the center of the chuck 103. Then, the chuck center position is regarded as the center of the table 105. It is to be noted that the center of the table 105 is used to calculate the position correction amount of the imaging location 201, which will be calculated later, and is not the center determined from the outer shape of the table.
[0019] The chuck 103 has a built-in temperature sensor 104, and the temperature of the chuck 103 can be constantly measured using the temperature sensor 104. The accuracy of the temperature measurement is particularly important here, and a temperature sensor using a platinum resistance thermometer is the most suitable. A typical temperature sensor using a platinum resistance thermometer uses a platinum wire manufactured so that the electrical resistance at 0°C is 100Ω, and the electrical resistance of the platinum wire is measured using a bridge circuit. This makes it possible to measure the electrical resistance of the platinum wire without being affected by the electrical resistance of the wiring between the sensor and the measuring instrument. Since the platinum wire is a stable material, the change in electrical resistance occurs only due to temperature change, and the relationship between temperature and electrical resistance is stable. Therefore, by referring to the correspondence between electrical resistance and temperature, highly accurate temperature measurement is possible. In addition, electrical resistance measurement has the characteristic of being resistant to electromagnetic noise compared to a method of measuring the voltage generated by a sensor.
[0020] The imaging device 100 mounts a sample 102 on a chuck 103 and images the sample 102 through a column 101. In the case of an inspection device used in the semiconductor manufacturing process, the sample 102 is a semiconductor wafer. In recent years, wafers with a diameter of 300 mm are often used, and the thickness of the wafer is generally less than 1 mm. In an apparatus that handles a 300 mm diameter wafer, the diameter of the chuck 103 is also close to 300 mm, but the thickness of the chuck 103 is designed to be sufficiently thicker than the wafer, so that the chuck 103 has high rigidity. In addition, the chuck 103 uses electrostatic force to strongly adsorb the wafer to the surface, making it possible to match the flatness of the wafer to that of the chuck 103. This makes it possible to eliminate the warp even if the wafer itself has a warp in the thickness direction, and to ensure a certain degree of flatness.
[0021] In an imaging device using an electron microscope, the column 101 corresponds to the electron microscope. The column 101 passes an electron beam through its center, so the position where the center of the column 101 is lowered becomes the imaging location 201. The column 101 is fixed inside the imaging device 100, and when the imaging location 201 of the sample 102 is to be moved significantly, the XY stage 108 moves to move the mounted sample 102. Furthermore, when a scanning electron microscope is used as the imaging device 100, the position where the electron beam is scanned can be changed to some extent. For this reason, when the imaging location is to be moved by a small amount, the XY stage 108 is fixed and the scanning position of the electron beam is changed.
[0022] When photographing a sample at high magnification, it is necessary to grasp the position of the photographing location 201 with high accuracy. For position measurement, a mirror and a laser interferometer installed on table 105 are used. Specifically, as shown in FIG. 2, an X-mirror 106 and a Y-mirror 110 are installed on table 105. Then, an X-axis laser interferometer 109 is provided facing the X-mirror 106, and a Y-axis laser interferometer 111 is provided facing the Y-mirror 110. The X-axis laser interferometer 109 and the Y-axis laser interferometer 111 are fixed to the photographing device 100.
[0023] The X-axis laser interferometer 109 performs highly accurate distance measurement in the X-axis direction by causing interference between the laser light irradiated to the X-mirror 106 and the laser light reflected back from the X-mirror 106, and counting the interference fringes. This allows a highly accurate measurement of a distance 202 between the X-axis laser interferometer 109 fixed to the imaging device 100 and the reflecting surface of the X-mirror 106. Similarly, the Y-mirror 110 and the Y-axis laser interferometer 111 allow a highly accurate measurement of a distance 203 between the Y-axis laser interferometer 111 and the reflecting surface of the Y-mirror 110. After the sample 102 is mounted on the chuck 103 and adsorbed, the relative distance between the sample 102 and the X-mirror 106 and the relative distance between the sample 102 and the Y-mirror 110 are fixed, and therefore the relative distance between the coordinates of the imaging location 201 and the X-mirror 106 and the relative distance between the coordinates of the imaging location 201 and the Y-mirror 110 are also fixed. Therefore, if the positions of X mirror 106 and Y mirror 110 are known with high accuracy, the coordinates of imaging location 201 on sample 102 can also be known with high accuracy.
[0024] The coordinate system expressing the positional deviation of the photographing location 201 described later is the XY coordinate system shown in Fig. 2, with the above-mentioned chuck center as the origin. In this coordinate system, the direction away from the mirror is the positive direction of the coordinate axis for each of the X and Y axes. Although there is some deviation between the center of the sample and the center of the chuck, they are assumed to be coincident here. In addition, since the position of the column 101 that photographs the sample is fixed, when the stage moves in the positive direction of the coordinate axis, the location of the sample that is photographed will be shifted in the negative direction of the coordinate axis.
[0025] The imaging device 100 includes a control unit, not shown in Figs. 1 and 2. The control unit acquires measurement results from the X-axis laser interferometer 109 and the Y-axis laser interferometer 111. Based on the measurement results of the X-axis laser interferometer 109 and the Y-axis laser interferometer 111, the position of the table 105 is grasped, and the center of the coordinate system, i.e., the center of the chuck 103 (=the center of the sample 102), is grasped. Then, the coordinates (X, Y) in the grasped coordinate system are set as the imaging location 201. Hereinafter, the coordinates of the imaging location 201 are set as (X, Y).
[0026] The control unit also acquires the chuck temperature acquired by the temperature sensor 104. The control unit then uses these to perform a position correction process, which will be described later. Thereafter, the control unit corrects the imaging location of the sample 102 by controlling the scanning position of the electron beam in the column 101, or by controlling a moving mechanism, such as a motor, that moves the XY stage 108. The control unit may also acquire an image captured by the column 101, and perform a position correction process using the image. The operation of the imaging device 100 will be described in detail below.
[0027] 3 shows an operation flow of the imaging device 100. Note that the sample 102 is placed on the chuck 103 at the start of the process flow shown in FIG.
[0028] <Step S1> The imaging device 100 performs alignment in step S1. Alignment is a process for accurately measuring the position of the sample 102 relative to the XY stage 108, and is a process for grasping the position of the sample 102 relative to the XY stage 108 based on the alignment mark formed on the sample 102. It is assumed that a circuit pattern has been formed on the sample 102 in a process prior to step S1, and that an alignment mark has been formed to serve as a coordinate marker. The column 101 adjusts the stage position so that the design coordinate of the alignment mark, that is, the position where the alignment mark should be, is the imaging position, and then images the location at a low magnification to image the alignment mark. Then, the control unit calculates the coordinate of the imaged alignment mark, and calculates the deviation between the calculated coordinate and the design coordinate of the alignment mark, thereby obtaining the position of the sample 102 relative to the XY stage 108. By photographing the alignment mark at multiple locations, it is possible to know the center misalignment when the sample 102 is mounted on the XY stage 108, the rotation angle, and the expansion / contraction state reflecting the sample temperature. In this way, the alignment process is a process of measuring the relative position of the sample 102 with respect to the XY stage 108 with high accuracy. The chuck 103 and the table 105 are fixed to the XY stage 108. Therefore, by measuring the relative position of the sample with respect to the XY stage 108 with high accuracy, the relative position of the sample 102 with respect to the chuck 103 and the table 105 can also be known. The relative position known by the alignment is used to determine the photographing location 201, so that the sample 102 can be photographed accurately.
[0029] For example, when it is desired to photograph a point of the sample 102 that is (X1, Y1) away from the sample center, if the chuck center and the sample center are coincident, the photographing location 201 may be specified as (X1, Y1) in the coordinate system shown in FIG. 2 with the chuck center as the origin. On the other hand, consider a case where the chuck center and the sample center are displaced by (X2, Y2) when the sample 102 is mounted on the chuck 103. In this case, since the chuck center and the sample center are displaced, it is not possible to photograph a point that is (X1, Y1) away from the sample center even if the photographing location 201 is specified as (X1, Y1) in the coordinate system shown in FIG. 2 with the chuck center as the origin. In such a case, by performing alignment, it is possible to know that the sample 102 was displaced by (X2, Y2) from the chuck 103 when it was mounted. 2, the imaging device 100 can accurately image a point that is (X1, Y1) away from the center of the sample by specifying the imaging location 201 as (X1-X2, Y1-Y2) in the coordinate system with the chuck center as the origin. In the following, the coordinates used are those corrected based on the relative positions of the chuck 103 and the sample 102 determined by alignment.
[0030] <Step S2> In step S2, the control unit records as initial temperatures the chuck temperature during alignment measured by temperature sensor 104 and the table temperature during alignment calculated based on the chuck temperature during alignment in a manner similar to step S4b described below. Note that the chuck temperature refers to the temperature of chuck 103, and the table temperature refers to the temperature of table 105.
[0031] <Step S3> In step S3, the control unit controls a moving mechanism for moving the XY stage 108 to move the XY stage 108 so that a portion of the sample 102 to be photographed, such as an inspection portion, becomes the photographing portion 201.
[0032] <Step S4> The photographing device 100 performs step S4 in parallel with step S3. In this embodiment, step S4 is performed in parallel with step S3, but it may be performed independently of step S3. For example, step S4 may be repeatedly performed at an independent cycle.
[0033] In step S4, the amount of position correction of the photographing location 201 is calculated. Specifically, the amount of position correction in the X-axis direction ψ is calculated using equations (1) and (2). X and the position correction amount in the Y-axis direction ψ Y Calculate.
[0034]
number
[0035]
number
[0036] Table 1 shows a list of variables used in the formulas in this specification.
[0037] [Table 1]
[0038] 4 shows a detailed flow of step S4 in the embodiment 1. Step S4 in the embodiment includes steps S4a to S4e.
[0039] <<Step S4a>> In step S4a, the temperature sensor 104 measures the chuck temperature.
[0040] <<Step S4b>> In step S4b, the control unit calculates the table temperature based on the chuck temperature. For example, the control unit calculates the table temperature based on the chuck temperature T chuck Using the time series data, the table temperature T table Calculate.
[0041]
number
[0042] <<Step S4c>> In step S4c, the control unit calculates the amount of change in the relative distance between the center of the sample and the mirror by calculating the amount of thermal expansion of table 105. Specifically, the amount of movement of X-mirror 106 in the X-axis direction relative to the center of the sample is calculated by multiplying the distance L from the center of table 105 to the reflecting surface of X-mirror 106 by the distance L x The linear expansion coefficient C2 of the table material and the temperature rise ΔT of the table table This is equivalent to finding the first term of equation (1). Similarly, find the amount of movement of the Y mirror 110 in the Y-axis direction relative to the center of the sample, that is, the first term of equation (2). Here, the amount of table temperature rise ΔT table is the difference between the latest calculated table temperature and the initial (alignment) calculated table temperature. The table center is considered to be the same as the chuck center. This is based on the fact that the chuck support pillars connecting the table and the chuck are arranged so that the center of gravity of the connection point coincides with the chuck center. Although the table center is originally defined as the same as the chuck center, it is considered that even if the two expand thermally at their respective temperatures, the center will continue to coincide even after the thermal expansion as long as the center of gravity of the chuck support pillar coincides with the chuck center. Step S4c makes it possible to grasp the amount of positional deviation of the photographing location 201 caused by the movement of the X mirror 106 and the Y mirror 110 due to the thermal expansion of the table 105, and the change in the distances 202 and 203 shown in FIG. 2. Note that when the temperature of the table 105 rises, the X mirror 106 moves in the negative direction of the X axis, and the Y mirror 110 moves in the negative direction of the Y axis, so that a minus sign is added before the coefficient C2 in Equation 1(1) and Equation (2). The table material may be a metal ceramic composite material, in which case the linear expansion coefficient C2 of the table material is 9.0×10 -6 It becomes.
[0043] <<Step S4d>> In step S4d, the amount of thermal expansion of the sample 102 is calculated. That is, in step S4d in this embodiment, the second term of formula (1) and the second term of formula (2) are obtained. In step S4d, the chuck temperature T chuck is considered to be equal to the sample temperature, and the chuck temperature T chuck The temperature difference between the initial temperature and the latest temperature ΔT chuck is regarded as the temperature rise of sample 102. Since normal thermal expansion occurs isotropically, the rate of increase obtained by dividing the increase in the distance between any two points by the original distance is proportional to the temperature rise, and the proportional coefficient at this time is the linear expansion coefficient of the material. Also, when thermal contraction occurs due to a decrease in temperature, it can be treated as negative thermal expansion in response to a negative temperature rise, so thermal contraction can also be treated as "thermal expansion".
[0044] In this embodiment, the fixed point of the sample 102 during thermal expansion (a point on the sample 102 whose position relative to the chuck 103 does not change) is the center of the sample 102, and isotropic radial thermal expansion occurs from the center of the sample 102. Furthermore, the center of the sample 102 and the center of the chuck 103 are assumed to coincide. In this case, in the coordinate system shown in FIG. 2, the X-direction displacement due to thermal expansion of the sample 102 is proportional to the X-coordinate of the photographed location 201, and the Y-direction displacement is proportional to the Y-coordinate of the photographed location 201. Therefore, in step S4d, the linear expansion coefficient C5 of the sample and the temperature rise ΔT are calculated based on the X-coordinate of the sample 102. chuck The amount multiplied by this is calculated as the amount of thermal expansion in the X-axis direction of sample 102, and the linear expansion coefficient C5 and the amount of temperature rise ΔT of the sample are calculated on the Y coordinate of sample 102. chuck The amount multiplied by is calculated as the amount of thermal expansion in the Y-axis direction of the sample 102. By step S4d, the amount of positional deviation of the photographing location 201 caused by the thermal expansion of the sample 102 can be grasped. The value of the linear expansion coefficient C5 of the sample is, for example, 3.9×10 -6 It becomes.
[0045] <<Step S4e>> In step S4e, the positional deviation amounts calculated in steps S4c and S4d are combined based on equations (1) and (2). In this manner, the positional deviation amount of the imaging point 201 due to thermal expansion is calculated as a position correction amount. X is the positive value in the direction approaching the X mirror, and the positive value in the Y axis direction is the position correction amount ψ Y is written with the direction approaching the Y mirror as positive.
[0046] <Step S5> In step S5 shown in Fig. 3, imaging is performed by shifting the imaging location 201 by the amount of position correction calculated in step S4. This allows position correction, making it possible to perform imaging with improved positional accuracy. Furthermore, while the time that has elapsed since alignment is short, the temperature change in the sample 102 and table 105 is small, so it is possible to predict in advance that positional deviation due to thermal expansion will be small, and imaging may be performed without position correction. Note that, if there are multiple locations to be imaged on the sample 102, steps S3 to S5 are repeated for each imaging location.
[0047] <Explanation of formula (3)> Here, the derivation of formula (3) will be explained. Formula (4) is a model of the heat transfer phenomenon by thermal conduction, in which the amount of heat transferred from table 105 to chuck 103 is taken as positive, and heat is considered to be transferred in proportion to the temperature difference between the two, with K being the proportionality coefficient.
[0048]
number
[0049] Equation (4) also holds when the table temperature is lower than the chuck temperature, in which case the amount of heat transferred from the table 105 to the chuck 103 is negative.
[0050] Equation (5) is a relational expression when the chuck temperature increases due to the amount of heat transferred from the table 105.
[0051]
number
[0052] In reality, heat is transferred to the chuck 103 through other routes, but a bold model is used to construct a practical relational equation. Equation (5) is derived from the law of conservation of energy and is a relational equation when the temperature of the chuck 103 rises due to the amount of heat acquired by the chuck 103. In other words, the time derivative (rate of change over time) of the chuck temperature multiplied by the mass of the chuck 103 and its specific heat is equal to the amount of heat acquired by the chuck 103. Since the actual chuck 103 is not made of a single material, the specific heat is not simply determined. In addition, the heat acquired by the chuck 103 is also transferred to the sample adsorbed by the chuck 103, so the mass and specific heat of the sample are also related here. Therefore, equation (5) is written in a conceptual sense.
[0053] From equations (4) and (5), the heat quantities refer to the same thing, so the relationship in equation (6) can be derived.
[0054]
number
[0055] This leads to equation (7), which is a conceptual equation for calculating the table temperature. Equation (4) models the amount of heat transferred from the table 105 to the chuck 103.
[0056]
number
[0057] Equation (3) is derived from equation (7) and is an equation for calculating the table temperature from the measured data of the chuck temperature. That is, it shows that the table temperature is calculated by multiplying the time derivative (time change rate) of the chuck temperature by the time constant C1 and adding the result to the chuck temperature. As shown in equation (7), the time constant C1 is formed by the mass, specific heat, and heat transfer coefficient of the chuck 103, but it is better to determine it in accordance with the actual behavior rather than calculating it from individual physical quantities. The time constant C1 is, for example, 2500 seconds. Note that time differentiation of the measured data of the chuck temperature is easily affected by noise, and the value is difficult to stabilize. Originally, in order to perform high-precision positioning, it is desirable that the stage temperature does not change as much as possible, and it is assumed that the chuck temperature and table temperature, which are the subject of this invention, also change only slightly. Therefore, it is very difficult to predict the table temperature with high precision using the time derivative of the measured data of the chuck temperature. Therefore, next, a means for stabilizing the time derivative will be described.
[0058] <<Method for stabilizing time derivatives>> When you have time series data, you can find the simple time derivative by dividing the difference from the data one step before by the time step at that time. However, because this calculation involves dividing a small quantity by a small quantity, even a small error can cause the value to change significantly, making it difficult to obtain a stable value. This is where techniques for smoothing the measured data come in handy. Equation (8) is the relational equation for finding the smoothed output variable y from the input variable u.
[0059]
number
[0060] Here, it is written in the form of a second-order differential equation of the variable y. The idea behind smoothing is as follows. It is assumed that the noise components are high-frequency components compared to the time change of the true physical quantity, and the true physical quantity is extracted by smoothing the data by removing the high-frequency components. Since equation (8) is a differential equation of the variable to be smoothed, the smoothed variable has a property that it is determined by integrating the input variable. Due to the existence of this integration process, the noise components that fluctuate positively / negatively in a short time are offset, and the physical quantity from which the noise has been removed is extracted. Although equation (8) is written in the form of an input variable u and a smoothed output variable y, when applied to the present invention, the measured data of the chuck temperature corresponds to the input variable u, and the output variable y corresponds to the smoothed chuck temperature. Since the third term on the left side of equation (8) and the first term on the right side are in an equal relationship, it can be seen that the gain of the output with respect to the input is 1. In other words, when the input data changes slowly, the value as it is becomes the output data. The differential terms of the first and second terms on the left side of equation (8) are defined by the angular frequency ω and the damping coefficient ζ. The angular frequency ω can be determined from the time constant τ for smoothing, as shown in equation (9).
[0061]
number
[0062] The meaning of the time constant τ is that noise components with a period greater than τ are removed by the smoothing process. However, a time lag occurs in the smoothed data compared to the original data, and this lag time is approximately the time constant τ. Therefore, if the time constant τ is set large, even noise with a large period can be removed and powerful smoothing can be achieved, but the time lag of the smoothed data will be large. On the other hand, if the time constant τ is set small in an attempt to reduce the time lag, the smoothing will be weak and the fluctuation of the time derivative value will be large. For this reason, it is necessary to determine the true time change rate of the physical quantity of interest and select the time constant τ appropriately.
[0063] The smaller the damping coefficient ζ is set to, the smaller the time lag associated with smoothing will be. On the other hand, it has the effect of amplifying the noise components that are synchronized with the angular frequency ω. Conversely, if the damping coefficient ζ is made larger, the time lag will increase, but the noise will not be amplified. The smallest ζ within the range where there is no noise amplification has been theoretically found to be ζ = 1 / √2 (≒ 0.707). For this reason, this value will be used for the damping coefficient ζ.
[0064] In solving equation (8), the first derivative of the output variable is newly defined as variable x, and using this, equation (8) is rewritten as simultaneous first-order differential equations (10) and (11) with two unknown variables.
[0065]
number
[0066]
number
[0067] In order to fit this equation to actual data, we discretize equations (12) to (14).
[0068]
number
[0069]
number
[0070]
number
[0071] Here, the subscripts of the variables indicate the time series number of the data, with the nth data being the most recent, and the n-1th being the data immediately preceding that. Equation (12) indicates the time interval of the data, but this time interval does not need to be constant, and the time interval may vary for each number. By expanding and solving equations (13) and (14), equations (15) and (16) are obtained.
[0072]
number
[0073]
number
[0074] Equation (15) shows that the latest output variable x is found using the latest input variable u and the previous output variables x and y. Furthermore, once the latest x has been found, the latest output y is found using equation (16). After this, when the next input is obtained, the previous most recent value becomes the previous value, and in combination with the latest input, the latest output is obtained. By repeating this process, it is possible to always obtain smoothed data for the latest measurement data and the time derivative of that smoothed data. Furthermore, it is possible to start the calculation by setting the initial values of the output variables as shown in equations (17) and (18).
[0075]
number
[0076]
number
[0077] Using the above technology, it is possible to obtain a stable time derivative from noisy measured temperature data, which makes it possible to practically apply equation (3) and grasp the temperature of the table that is not measured with high accuracy, thereby making it possible to calculate the amount of thermal expansion with a practical level of accuracy.
[0078] As described above, in this embodiment, it was determined that the thermal expansion of the sample and the thermal expansion of the table on which the mirror is mounted are the main factors, and a means for quantitatively predicting each factor was constructed. Based on this, the temperature of the sample was grasped using the measurement value of the temperature sensor of the chuck on which the sample is mounted, and the temperature of the table was grasped by processing and calculating the measurement value of the temperature sensor of the chuck. When the temperatures of two locations are grasped from the measurement data of one location, the degree of freedom for calculating the thermal expansion is doubled, and it becomes possible to calculate the thermal expansion with higher accuracy. By predicting the positional deviation of the photographing location 201 based on the amount of thermal expansion of these two points and correcting the position by the predicted amount, it becomes possible to greatly improve the positional accuracy of the location where the sample is photographed.
[0079] Therefore, according to this embodiment, the sample can be photographed with high accuracy regardless of the presence or absence of thermal expansion of the sample. As a result, the positional deviation between the target location and the location actually photographed is reduced, and even when photographing at high magnification, the location to be inspected can be photographed accurately, enabling inspection work to be performed with high throughput. EXAMPLES
[0080] 5 shows a detailed flow of step S4 in the embodiment 2. Most of the process is the same as in the embodiment 1, so only the differences will be explained here.
[0081] In step S4 of this embodiment, a calculation of the rotation angle of the mirror is added in step S4f, and the position correction amount ψ in the X-axis direction is calculated using equations (19) and (20). X and the position correction amount in the Y-axis direction ψ Y Calculate.
[0082]
number
[0083]
number
[0084] When the latest table temperature is calculated in step S4b, the amount of thermal expansion of the table 105 is calculated in step S4c to calculate the corresponding movement amount of the mirrors 106 and 110. However, the thermal deformation of the table 105 may be deformation other than isotropic thermal expansion. That is, in isotropic thermal expansion, the shape of the table 105 after thermal expansion is similar to the shape of the table before expansion. If the original shape of the table 105 is square, in isotropic thermal expansion, the square shape is maintained even after expansion. However, since the actual table 105 is an integral structure with the XY stage 108 present below it, it is connected to parts made of various materials, and the temperature distribution occurs, so that the temperature rise amount is not uniform, and therefore it does not necessarily become isotropic thermal expansion. For this reason, even if the shape of the table 105 viewed from above was originally square, the table 105 after thermal deformation may become a trapezoid, a diamond, or another shape. In that case, a rotational displacement is added to the displacement due to expansion for mirrors 106, 110 placed on table 105. For such mirror displacement, the displacement distribution in the X direction can be expressed by the displacement in which the center of X mirror 106 moves in the X direction and the rotation angle around the center of X mirror 106. Similarly, for Y mirror 110, the displacement distribution in the Y direction can be expressed by the displacement in which the center of Y mirror 110 moves in the Y direction and the rotation angle around the center of Y mirror 110.
[0085] The driving force behind the table shape no longer being similar to its original shape due to temperature changes is thermal expansion, which is subject to various constraints and results in a change in shape. Therefore, the deformation that causes the mirror angle to change can also be considered to be driven by the temperature rise of the table. Therefore, the mirror rotation angle is calculated in step S4f based on the table temperature calculated in step S4b.
[0086] The coefficient C3 in equation (19) is the temperature coefficient of the rotation angle of the X-mirror 106, and is a proportionality coefficient assuming that the rotation angle increases in proportion to the increase in the table temperature. The coefficient C3 is, for example, 0.2×10 -6 Here, the rotation of X mirror 106 is such that the mirror center is set to Y=0, and this mirror center is used as the reference for rotation, so that the relative distance between the imaging location at Y=0 and the mirror does not change even if X mirror 106 rotates. In addition, since the rotation angle here is a very small amount, the length of the arc due to the rotational displacement and the displacement in the X direction can be considered to be the same. Therefore, the rotation angle (unit: radian) of X mirror 106 is calculated by multiplying the temperature rise of table 105 by coefficient C3, and the displacement in the X direction can be calculated by multiplying this by the Y coordinate equivalent to the radius. Coefficient C4 in equation (20) also works in the same way as C3, and the proportionality coefficient when Y mirror 110 rotates in proportion to the rise in table temperature is set to C4, and the rotation angle (unit: radian) of Y mirror 110 is obtained by multiplying C4 by the rise in table temperature, and the displacement in the Y direction is obtained by multiplying this by the X coordinate. Coefficient C4 is, for example, -0.3×10 -6 It becomes.
[0087] That is, in step S4f, the Y coordinate of the sample 102 is calculated by dividing the temperature coefficient C3 of the rotation angle of the X mirror 106 and the table temperature change ΔT table Multiplying by C3 Y ΔT in equation (19) table In addition, the temperature coefficient C4 of the rotation angle of the Y mirror 110 and the table temperature change ΔT table Multiplying by C4·X·ΔT in equation (20) tableThe other terms in formula (19) are the same as those in formula (1), and the other terms in formula (20) are the same as those in formula (2). Then, in step S4e, the positional deviations calculated in steps S4c, S4d, and S4f are combined based on formulas (19) and (20) to obtain a positional correction amount ψ in the X-axis direction. X and the position correction amount in the Y-axis direction ψ Y Calculate.
[0088] According to this embodiment, even if the table 105 undergoes thermal expansion and the mirror rotates, the sample can be photographed with high accuracy. EXAMPLES
[0089] Fig. 6 shows an operation flow of the imaging device 100 in the embodiment 3. Fig. 7 shows a detailed flow of step S4 in the embodiment 3. Here, only the points different from the embodiment 2 will be explained.
[0090] In the first and second embodiments, it was assumed that when the temperature of the sample rises and the sample expands thermally, the center of the sample becomes a fixed point and expands isotropically. This holds true when the force of the chuck 103 to attract the sample 102 acts on the entire surface of the chuck and a uniform frictional force is generated. That is, since the chuck 103 and the sample 102 are made of different materials, their linear expansion coefficients are different, and even if the two maintain the same temperature and rise in temperature, their thermal expansion amounts will be different. In order to compensate for this difference in thermal expansion amount, it is necessary for slippage to occur between the chuck 103 and the sample 102. A frictional force based on the attraction force is generated against the sliding motion at this time. For this reason, the sliding motion of the sample 102 should occur so that the energy consumption due to the frictional force is minimized. The relative slippage between the sample 102 and the chuck 103 is minimized when the center of the sample 102 becomes a fixed point and slides radially from there. Therefore, when the frictional force acts uniformly on the chuck 103, the center of the sample 102 should be a fixed point. However, in reality, the adhesion force between the sample 102 and the chuck 103 is not uniform, and the friction coefficient is not necessarily constant. Therefore, slippage may occur between the sample 102 and the chuck 103, and a position different from the center of the chuck may become a fixed point. Taking this into consideration, the amount of positional deviation is calculated in Example 3.
[0091] In order to estimate the fixed point of the sample slippage in step S4, the imaging device 100 of this embodiment performs step S6 after step S5, as shown in FIG. 6. In step S6, the control unit performs image analysis on the image captured in step S5, and analyzes the actual positional deviation amount of the imaging location 201 as the actual position correction amount of the imaging location 201. This image analysis may be performed based on the circuit pattern or alignment marks formed on the sample 102. Also, the actual position correction amount of the imaging location 201 may be analyzed from multiple images obtained by capturing multiple locations. The analysis result of step S6 is used in the next step S4.
[0092] In addition, in step S4 of this embodiment, steps S4g and S4h are added as shown in FIG. 7, and the position correction amount ψ in the X-axis direction is calculated using equations (21) and (22). X and the position correction amount in the Y-axis direction ψY Calculate.
[0093]
number
[0094]
number
[0095] <Step S4g> In step S4g, the coordinates of the fixed point (X f ,Y f ) is estimated. Equations (23) to (30) show the procedure for obtaining the fixed-point coordinates when the sample slides on the chuck. In performing this task, we first use equations (23) and (24) to calculate the position correction amount of the shooting location 201 when the fixed-point coordinates are at the center of the sample, and this is called the advance predicted value.
[0096]
number
[0097]
number
[0098] When the sample 102 is actually photographed, the amount of positional deviation between the photographing position assumed from the stage position and the actually photographed position can be evaluated. Therefore, the difference between the predicted value of the position correction amount of the photographing location 201 and the position correction amount obtained by actual photographing is defined as the error of the predicted value by using the formulas (25) and (26). Note that the position correction amount ψ obtained by actual photographing is x real 、 ψ y real The result calculated in the previous step S6 is used for the position correction amount ψ x real 、 ψ y realWhen calculating, for example, the position correction amount ψ x real 、 ψ y real may be calculated and the average value of these may be used.
[0099]
number
[0100]
number
[0101] Comparing this with equations (23) and (24), we obtain equations (27) and (28) for calculating the fixed-point coordinates. The fixed-point coordinates obtained by these equations will be referred to as the prior values of the fixed-point coordinates for the time being.
[0102]
number
[0103]
number
[0104] Since equations (27) and (28) are calculated by dividing a small value by a small value, the values tend to become unstable. For this reason, we perform the operations in equations (29) and (30) to stabilize the values.
[0105]
number
[0106]
number
[0107] The left-hand side of equations (29) and (30) is the fixed-point coordinate resulting from stabilization. The first item on the right-hand side is the latest preliminary value of the fixed-point coordinate, and the other is the stabilized value of the previous fixed-point coordinate. The stabilization number m is set to a number greater than 1. For example, when m = 5, when the latest preliminary value of the fixed-point coordinate is obtained, it is diluted to 1 / 5 and used to update the fixed-point coordinate. By implementing such an operation, it is possible to prevent the fixed-point coordinate from changing suddenly.
[0108] The fixed point coordinates (X f ,Y f ) shall use the previous value and its initial value shall be 0.
[0109] <Step S4h> In step S4h, the amount of slippage of the sample is calculated. Specifically, the X-coordinate X of the fixed point estimated in step S4g is calculated. f The difference in linear expansion coefficient C6 between the chuck 103 and the specimen 102 and the change in chuck temperature ΔT chuck In addition, the Y coordinate Y of the fixed point estimated in step S4g is multiplied by f The difference in linear expansion coefficient C6 between the chuck 103 and the specimen 102 and the change in chuck temperature ΔT chuck Calculate the third term of equation (22) by multiplying by
[0110] The third term in equations (21) and (22) is a term related to the fixed point coordinates when the sample 102 slides on the chuck 103 due to the difference in the amount of thermal expansion between the sample 102 and the chuck 103. The XY coordinates of this fixed point are f and Y f Therefore, if the fixed point is the center of the sample, X f and Y fbecomes 0, and this term becomes 0. If the fixed point is different from the center of the sample, the fixed point moves in proportion to the fixed-point coordinates by the amount of thermal expansion of the chuck 103. However, when this expansion occurs, the sample 102 also expands, so the effective amount of movement is the amount of thermal expansion of the chuck 103 offset by the amount of thermal expansion of the sample 102. Therefore, the proportionality coefficient C6 of the amount of movement in proportion to the fixed-point coordinates is assigned the difference between the linear expansion coefficient of the chuck material and the linear expansion coefficient of the sample material. The proportionality coefficient C6 can be calculated, for example, as 3.2×10 -6 It becomes.
[0111] <Step S4e> In step S4e, the positional deviation amounts calculated in steps S4c, S4d, S4f, and S4h are combined based on equations (21) and (22). Note that equations (31) and (32) can be derived by applying equations (25) and (26) to the first and second terms of equations (21) and (22). Equations (31) and (32) may also be used in step S4e.
[0112]
number
[0113]
number
[0114] According to this embodiment, even if the sample slips due to thermal expansion, the sample can be photographed with high accuracy. EXAMPLES
[0115] 4 shows a detailed flow of step S4 in the fourth embodiment. Only the points different from the third embodiment will be described here.
[0116] The imaging device 100 of this embodiment includes a temperature sensor for measuring the table temperature on the table 105. Then, as shown in Fig. 9, the table temperature is measured in step S4i, and in steps S4c and S4f, the amount of thermal expansion of the table 105 and the rotation angle of the mirror are calculated using the table temperature measured in step S4i.
[0117] According to this embodiment, as in the third embodiment, it is possible to identify the fixed-point coordinates when the sample slides on the chuck, and then predict the amount of position correction of the imaging location with high accuracy. This makes it possible to image the sample with high accuracy by correcting the position. [Explanation of symbols]
[0118] 100 Imaging Device 101 Column 102 Sample 103 Chuck 104 Temperature Sensor 105 Table 106X Mirror 107 Chuck support 108 XY Stage 109 X-axis laser interferometer 110 Y Mirror 111 Y-axis laser interferometer 201 Shooting Location 202 X-axis stage position measurement distance 203 Y-axis stage position measurement distance
Claims
1. A table and a chuck whose back surface is fixed to the table and whose front surface adsorbs the sample; a temperature sensor for acquiring a temperature of the chuck; a column for imaging the sample mounted on the chuck; A mirror provided on the table; a laser interferometer that irradiates the mirror with a laser beam and measures the distance to the mirror; A control unit that calculates a position correction amount of an imaging location based on the temperature of the chuck, The control unit is calculating a temperature of the table based on the temperature of the chuck; Calculating the amount of thermal expansion of the table based on the temperature of the table; Calculating the thermal expansion of the sample based on the temperature of the chuck; A position correction amount for the imaging location is calculated based on the thermal expansion amount of the table and the thermal expansion amount of the sample. An imaging device characterized by:
2. The imaging device according to claim 1 , The control unit is Calculating a rotation angle of the mirror based on the temperature of the table; The position correction amount is calculated based on the thermal expansion amount of the table, the thermal expansion amount of the sample, and the rotation angle of the mirror. An imaging device characterized by:
3. The imaging device according to claim 1 , The control unit is Analyze the position correction amount of the photographed portion from the photographed image; estimating a fixed point on the specimen whose position relative to the chuck does not change based on the analytical position correction amount and the calculated position correction amount; calculating an amount of slippage of the specimen relative to the chuck based on the fixed point and a temperature of the chuck; The position correction amount is calculated based on the thermal expansion amount of the table, the thermal expansion amount of the sample, and the sliding amount of the sample. An imaging device characterized by:
4. The imaging device according to claim 3, The control unit analyzes a position correction amount of each of the plurality of photographed locations from a plurality of images obtained by photographing the plurality of locations in the analysis. An imaging device characterized by:
5. The imaging device according to claim 1 , The control unit is Calculating the temperature of the table based on the temperature of the chuck and the rate of change of the temperature of the chuck over time. An imaging device characterized by:
6. The imaging device according to claim 1 , A stage for moving the table is provided, The stage corrects the position of the photographing location in advance based on the position correction amount of the photographing location. An imaging device characterized by:
7. 1. A method for photographing a sample mounted on a chuck, comprising: a table temperature calculation step of calculating a temperature of a table to which the chuck is fixed based on the temperature of the chuck; a table thermal expansion amount calculation step of calculating a thermal expansion amount of the table based on the temperature of the table; a sample thermal expansion amount calculation step of calculating a thermal expansion amount of the sample based on the temperature of the chuck; a position correction amount calculation step of calculating a position correction amount of an imaging location based on the thermal expansion amount of the table and the thermal expansion amount of the sample; a position correction step of correcting the position of the photographing location based on the position correction amount; A photographing method comprising:
Citation Information
Patent Citations
Stage device, aligner, method for manufacturing semiconductor device, semiconductor manufacturing plant, and maintenance method for the aligner
JP2002118050A
Exposure equipment, exposure method and device production method
JP2017129875A
