A scanning method, apparatus, semiconductor device, medium, and program product
By generating first and second compensation data for closed-loop compensation, the alignment accuracy problem between the wafer and the mask caused by thermal effects and process is solved, thereby improving the accuracy and reliability of defect detection.
Patent Information
- Application Number
- CN202611327189.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-28
- Publication Date
- 2026-11-06
AI Technical Summary
In existing technologies, wafer expansion and deformation caused by thermal effects and manufacturing processes affect the alignment accuracy between DIEs, resulting in low defect detection accuracy.
The scanning trajectory is compensated by generating first compensation data, and the coordinate deviation of adjacent bare dies or bare die patterns is acquired in real time during the scanning process to generate second compensation data. The first and second compensation data are combined to perform closed-loop compensation, and the third compensation data is determined to improve the scanning accuracy.
It achieves higher alignment accuracy in wafer and mask inspection, reduces false defect signals and missed detections, and improves the accuracy and reliability of inspection.
Smart Images
Figure CN123046091A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of defect detection technology, and in particular to a scanning method, apparatus, semiconductor device, medium, and program product. Background Technology
[0002] Bright field inspection (BFI), dark field inspection (DFI), or pattern mask inspection (PMI) detects defects by comparing the bare die (DIE) with another DIE. Therefore, the higher the alignment accuracy between the two DIEs, the better. Wafer expansion due to thermal effects or manufacturing process deviations can affect the alignment accuracy between DIEs. To address wafer deformation, a feedforward compensation (premap) scheme is typically used. Feedforward premap is an n... The compensation matrix of m is equivalent to cutting a wafer into n parts. The function uses m small rectangles, each with a compensation value (premap value). However, even after applying feedforward compensation, the defect detection accuracy remains poor.
[0003] Therefore, improving the accuracy of defect detection is a problem that needs to be solved in this field. Summary of the Invention
[0004] This application discloses a scanning method, apparatus, semiconductor device, medium, and program product to solve the problem of low accuracy in defect detection.
[0005] In a first aspect, this application provides a scanning method, comprising: performing coordinate compensation on the scanning trajectory of the object to be scanned using first compensation data corresponding to the object to be scanned; the object to be scanned is a wafer and / or a mask, the first compensation data is the deviation data between the ideal coordinates and the actual coordinates of a marker line in the object to be scanned, the marker line is a marker line set in a preset area of the object to be scanned, and the scanning trajectory includes multiple scanning lines; generating second compensation data; the second compensation data includes the coordinate deviation between two adjacent areas to be scanned located in the current scanning line during the scanning process of the scanning trajectory, the area to be scanned is the area of the bare die corresponding to the wafer and / or the area of the bare die pattern corresponding to the mask; determining third compensation data corresponding to subsequent scanning lines using the second compensation data and the first compensation data corresponding to the current scanning line, and performing a scanning operation on the subsequent scanning lines based on the third compensation data.
[0006] In this application, the first compensation data is essentially a static deformation mapping of the wafer and / or mask in a non-working state, reflecting the global and systematic spatial offset of the wafer and / or mask caused by thermal history or process stress. Before the wafer and / or mask scanning begins, this solution first uses this data to perform coordinate compensation on the scanning trajectory, essentially establishing a relatively accurate initial alignment reference for the entire scanning process. This provides a high-precision starting position for subsequent scanning, avoiding subsequent compensation failure due to excessive initial deviation. During the wafer and / or mask scanning process, the high-speed movement of the motion platform, continuous heat accumulation, and the non-uniformity of local material properties can all induce real-time deformation of the wafer and / or mask. This deformation cannot be fully predicted through static calibration before scanning. This solution generates second compensation data by obtaining the coordinate deviation of adjacent dies (patterns) in the current scanning row, essentially using the relative positional deviation between adjacent dies (patterns) as a real-time feedback signal. Because this signal originates from real-time measurements at adjacent positions within the same scan line, it possesses extremely high timeliness and local realism, accurately capturing transient errors caused by temperature drift, vibration, or other random disturbances during the scanning process. During scanning, the first compensation data serves as the basic reference, and the real-time deviation represented by the second compensation data is superimposed to determine the compensation value used for subsequent scan lines (especially those in the same direction), i.e., the third compensation value. This continuous updating of the accuracy of subsequent scan lines enables closed-loop compensation combining the first and second compensation data, significantly improving the overall alignment accuracy of the die (pattern) position. This provides a higher-quality scanning reference for subsequent defect detection (such as bright-field inspection, BFI), effectively suppressing false defect signals or missed defects caused by positioning deviations, ultimately ensuring the accuracy and reliability of wafer and / or mask inspection.
[0007] In one possible embodiment, determining the third compensation data corresponding to the subsequent scan line using the second compensation data and the first compensation data corresponding to the current scan line includes: determining the next scan line in the same direction as the current scan line as the target scan line; updating the first compensation data corresponding to the target scan line based on the second compensation data corresponding to the current scan line to obtain the third compensation data corresponding to the target scan line.
[0008] In this embodiment, considering the trajectory direction of different scanning operations, the force and motion law of the device in the same direction of scanning are basically the same, and the coordinate deviation change trend of adjacent rows is also related. Here, the first compensation data corresponding to the same direction of scanning operation can be updated based on the real-time second compensation data, which can improve the compensation accuracy and avoid the error correction caused by the different motion laws of opposite direction scanning.
[0009] In one possible implementation, updating the first compensation data corresponding to the target scan line based on the second compensation data corresponding to the current scan line to obtain the third compensation data corresponding to the target scan line includes: determining a current interval matching relationship between a second interval corresponding to the second compensation data and a first interval of the first compensation data corresponding to the target scan line; the second interval includes the region of one or more dies corresponding to the wafer and / or the region of one or more dies corresponding to the mask, and the first interval is the region of one or more dies corresponding to the wafer and / or the region of one or more dies corresponding to the mask; updating the first compensation data corresponding to the target scan line based on the current interval matching relationship using the second compensation data to obtain the third compensation data corresponding to the target scan line.
[0010] In this embodiment, the actual wafer is composed of a large number of independent bare dies, and the mask is composed of a large number of independent bare die patterns. The manufacturing deviations and equipment processing errors of different bare die regions or bare die pattern regions are not uniform, and the accuracy of the second compensation data and the first compensation data may also be different. Here, the effective range of the second compensation data on the wafer and / or mask is mapped to the relevant first compensation data. Taking the real-time second compensation data as the standard can improve the accuracy of the final third compensation data.
[0011] In one possible implementation, updating the first compensation data corresponding to the target scan line using the second compensation data based on the current interval matching relationship to obtain the third compensation data corresponding to the target scan line includes: calculating the deviation between the second compensation data and the first compensation data corresponding to the same-direction scan lines that have already undergone scanning operations, and accumulating the corresponding number of deviation calculation results to obtain a total deviation result; updating the first compensation data corresponding to the target scan line using the total deviation result to obtain the third compensation data corresponding to the target scan line.
[0012] In this embodiment, for the compensation data of each scan line, the total deviation result of the compensation data corresponding to the previous scan lines is calculated by accumulation, and the first compensation data corresponding to the target scan line is updated using the total deviation result. The resulting third compensation data includes the real-time deviation and the cumulative deviation, and the coordinates of the final target scan line will be more accurate. This avoids the situation where the overall scanning positioning error increases as the scan line progresses, and improves the compensation accuracy.
[0013] In one possible implementation, before determining the third compensation data corresponding to subsequent scan rows using the second compensation data and the first compensation data corresponding to the current scan row, the method further includes: calculating the mean and standard deviation of a plurality of second compensation data corresponding to each of the scanned regions in the current scan row; if the difference between a single second compensation data corresponding to the current scan row and the mean is not less than a deviation threshold, then the corresponding single second compensation data is removed, and finally a new second compensation data corresponding to the current scan row is obtained; the deviation threshold is a threshold calculated based on the standard deviation; determining the third compensation data corresponding to subsequent scan rows using the second compensation data and the first compensation data corresponding to the current scan row includes: determining the third compensation data corresponding to subsequent scan rows using the new second compensation data and the first compensation data corresponding to the current scan row.
[0014] In this embodiment, outliers in the second compensation data are filtered out by standard deviation calculation to ensure the validity of the second compensation data and improve the compensation accuracy.
[0015] In one possible implementation, the process of generating the first compensation data includes: scanning the marker lines of the object to be scanned, and obtaining deviation data corresponding to the marker lines based on the deviation between the actual coordinates of the scanned bare die and / or bare die pattern and the ideal coordinates corresponding to the marker lines; and calculating the compensation data corresponding to the object to be scanned using the deviation data through linear interpolation to obtain the first compensation data corresponding to the object to be scanned.
[0016] In this embodiment, it is not necessary to scan every single die and / or die pattern to obtain the deviation data of the entire wafer and / or mask, which greatly reduces the initial measurement time and ensures production efficiency. At the same time, linear interpolation can fit the overall systematic deviation of the wafer, which can reduce the pressure of subsequent data correction.
[0017] In one possible implementation, the process of generating the second compensation data includes: during the scanning process of the scanning trajectory, determining the position data of the area to be scanned in the current scan line obtained by scanning through a preset number of sub-scanning bands; determining the coordinate deviation between two adjacent areas to be scanned in the current scan line based on the position data; merging the coordinate deviations corresponding to the adjacent sub-scanning bands, and performing weighted calculation on the corresponding merged data to obtain the second compensation data corresponding to the current scan line.
[0018] In this embodiment, a full-line scan with a large width can be divided into multiple sub-scanning bands for parallel acquisition, which greatly shortens the acquisition time of a single scan line. This can adapt to the high-speed scanning requirements of large-size wafers and / or photomasks. At the same time, with the addition of weighted calculation, reasonable weights can be assigned to regions of different bare dies and / or regions of bare die patterns to avoid mutual influence of cross-regional deviations.
[0019] Secondly, this application provides a scanning apparatus for implementing the scanning method described above.
[0020] Thirdly, this application provides a semiconductor device for implementing the scanning method described above.
[0021] In one possible implementation, the semiconductor device includes a line scan device, a real-time alignment card, and a synchronization board; wherein, The line scanning device is used to perform scanning operations on the object to be scanned; The real-time alignment card is used to send the data obtained by the line scanning device through the scanning operation to the synchronization board. The synchronization board is used to perform the scanning method described above based on the data.
[0022] Fourthly, this application provides a computer-readable storage medium for storing a computer program, which, when executed by a processor, implements the scanning method described above.
[0023] Fifthly, this application provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the scanning method described above. Attached Figure Description
[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0025] Figure 1 A physical structure diagram of a bright-field defect detection machine provided in an embodiment of this application; Figure 2 A schematic diagram of a wafer scanning area provided in an embodiment of this application; Figure 3 This is another schematic diagram of a wafer scanning area provided in an embodiment of this application; Figure 4 A schematic diagram of feedforward compensation provided for an embodiment of this application; Figure 5A flowchart of a defect detection method provided in an embodiment of this application; Figure 6 A flowchart of a first compensation data update provided in an embodiment of this application; Figure 7 A specific flowchart of the first compensation data update is provided for an embodiment of this application; Figure 8 A flowchart of a second compensation data generation method provided in this application embodiment; Figure 9 A schematic diagram illustrating the interaction data between a synchronization board and an RTA card, provided as an embodiment of this application; Figure 10 A flowchart illustrating a specific defect detection method provided in this application embodiment; Figure 11 This is a schematic diagram of an x-axis offset provided in an embodiment of this application; Figure 12 This application provides a schematic diagram of the positional deviation of different DIEs in an embodiment. Figure 13 This is a schematic diagram of a position deviation update provided in an embodiment of this application; Figure 14 This is a schematic diagram of a semiconductor device structure provided in an embodiment of this application. Detailed Implementation
[0026] The technical solution of this application can be applied to bright field defect detection (BFI) equipment, as well as dark field defect detection (DFI), mask defect detection (PMI), and other scenarios involving scanning and detection of the object to be scanned. Specifically, taking a wafer as the object to be scanned and bright field defect detection as an example, bright field defect detection detects defects based on comparing two wafers (DIEs); for example... Figure 1The diagram shows the physical structure of the bright-field defect inspection machine. The host computer is connected to the MCS (Master Control Station) main control board, which in turn connects to a synchronization board. The synchronization board includes a synchronization card (such as Hi1213 for clock synchronization) and a Field Programmable Gate Array (FPGA). The stage is connected to the main control board to assist in wafer scanning and interacts with the FPGA via a probe card (PEG) signal. The FPGA can acquire wafer-related information collected by the grating ruler sensor (ABZ grating ruler). Specifically, the host computer can generate wafer layout information based on the planned scanning trajectory and send it to the main control board. The main control board sends wafer layout information to the motion stage and synchronization card respectively. The synchronization card sends the current swath (which can be understood as the information of the marker lines to be scanned) of the wafer to be scanned to the FPGA via DMA (Direct Market Access). Upon receiving the probe card signal, the FPGA starts scanning and outputs a TDI trigger pulse. When one wafer scan line is completed, an interrupt is generated and the synchronization card is notified. Upon receiving the interrupt, the synchronization card sends the next swath information to the FPGA to continue wafer scanning. The wafer scan area is as follows: Figure 2 As shown, the scanning process proceeds line by line in a serpentine (back-and-forth) scanning manner. Scanning begins on the left side of swath1, triggering the PEG_start signal. Scanning proceeds sequentially from DIE0 to DIE5, obtaining the corresponding PEG_end signals. Then, the transition is performed to configure the next swath information, i.e., scanning swath2. Therefore, scan lines can be divided into unidirectional scan lines (e.g., lines 1 and 3 are both scanned from left to right) and unidirectional scan lines (e.g., lines 1 and 2 are in opposite directions). Further, the wafer mark lines in the wafer scan area are as follows... Figure 3 As shown, the distance between DIEs can be set to 25.95 mm; correspondingly, in the wafer scan feedforward compensation (premap) scheme, the feedforward premap is an n The compensation matrix of m is equivalent to cutting a wafer into n parts. Each of the m-sized rectangles has a compensation value; for example... Figure 4 The diagram shown is a schematic of x-axis feedforward compensation.
[0027] Example 1
[0028] like Figure 5 As shown, this embodiment discloses a defect detection method, specifically including: Step S11: Use the first compensation data corresponding to the object to be scanned to perform coordinate compensation on the scanning trajectory of the object to be scanned.
[0029] In this application, the first compensation data is the deviation data between the ideal coordinates and the actual coordinates of the marker lines in the object to be scanned. The marker lines are several marker lines set in a preset area of the object to be scanned, for example, three marker lines in a fixed-size rectangle at a corner of the object. The first compensation data is a coordinate deviation value, such as an x-axis deviation of +0.2 (in length units). Essentially, the first compensation data is a static deformation mapping of the object to be scanned in a non-working state, reflecting the global and systematic spatial offset of the object due to thermal history or process stress.
[0030] Specifically, before scanning begins, the marker lines of the object to be scanned can be scanned first to obtain the actual coordinates of the marker points; at the same time, the ideal coordinates corresponding to the marker points can be obtained, and the deviation between the actual coordinates and the ideal coordinates of each marker point can be calculated to obtain the deviation data corresponding to the marker lines; then, based on the deviation data, first compensation data covering the entire object to be scanned can be generated through interpolation and other methods, and the first compensation data can be used to perform coordinate compensation on the scanning trajectory of the object to be scanned.
[0031] Understandably, before the scanning of the object to be scanned is started, this application first uses the first compensation data to perform coordinate compensation on the scanning trajectory, which is equivalent to establishing a relatively accurate initial alignment reference for the entire scanning process. This can provide a high-precision starting position for subsequent scanning and avoid subsequent compensation failure due to excessive initial deviation.
[0032] In one possible embodiment, the first compensation data can be generated as follows: The marker lines of the object to be scanned are scanned, and the deviation data corresponding to the marker lines is obtained based on the deviation between the actual coordinates of the scanned area and the ideal coordinates corresponding to the marker lines. The first compensation data is then obtained by calculating the compensation data corresponding to the object to be scanned using linear interpolation. For example, the marker lines of the wafer can be scanned in advance, and the measured coordinates of the marker points can be obtained according to the wafer coordinate system. Simultaneously, the offset of each marker point is calculated based on its theoretical coordinates. Then, based on the position corresponding to the marker points, linear interpolation is performed on the scanning direction (e.g., the x-axis) (linear extrapolation is used outside the range on both sides of the marker point, and linear interpolation is used between two marker points) to determine the initial feedforward compensation (premap) data. The feedforward compensation for the remaining rows is extended upwards and downwards through this map, maintaining consistency in each column, thereby obtaining the first compensation data covering the entire wafer. This eliminates the need to scan every bare die to obtain the deviation data for the entire wafer and / or mask, significantly reducing the initial measurement time and ensuring production efficiency. Simultaneously, linear interpolation can fit the overall system deviation of the wafer, reducing the pressure of subsequent data correction.
[0033] Step S12: Generate the second compensation data.
[0034] The second compensation data includes the coordinate deviation between two adjacent areas to be scanned in the current scan line during the scanning process of the scan trajectory. The areas to be scanned are the areas of the bare die corresponding to the wafer and / or the areas of the bare die pattern corresponding to the mask. The second compensation data refers to the coordinate deviation data at the same feature position between adjacent bare dies or bare die patterns on the same scan line, which is fed back in real time during the scanning process of the current scan line.
[0035] In this embodiment, during the scanning process, the high-speed movement of the stage, continuous heat accumulation, and the non-uniformity of local material properties may all induce real-time deformation of the wafer, which cannot be fully predicted by static calibration before scanning. Therefore, this application acquires the coordinate deviation between adjacent bare dies or between bare die patterns in real time as second compensation data during the scanning process of the current scan line.
[0036] In one possible embodiment, the second compensation data may come from feedback data from a real-time alignment card (RTA card). RTA feedback data can be defined as ΔX and ΔY, representing the deviation of a position within a DIE relative to the same position in the previous DIE. Because this signal originates from real-time measurements at adjacent positions within the same scan line, it possesses extremely high timeliness and local accuracy, enabling precise capture of transient errors caused by temperature drift, vibration, or other random disturbances during the scanning process.
[0037] It is understood that this application generates second compensation data by obtaining the coordinate deviation of adjacent dies or dies in the current scan line. Essentially, it uses the relative position deviation between adjacent dies or dies as a real-time feedback signal, thereby providing a real-time, local deviation basis for subsequent closed-loop compensation.
[0038] Step S13: Using the second compensation data and the first compensation data corresponding to the current scan line, determine the third compensation data corresponding to the subsequent scan line, and perform a scan operation on the subsequent scan line based on the third compensation data.
[0039] In this embodiment, during the scanning process, the first compensation data is used as the basic reference, and the real-time deviation represented by the second compensation data is superimposed to determine the compensation value for subsequent scan lines (especially unidirectional scan lines), namely the third compensation data, and the scanning operation is performed on the subsequent scan lines based on the third compensation data.
[0040] Specifically, before determining the third compensation data, the second compensation data corresponding to the current scan line can be processed to improve the effectiveness of the second compensation data, and then the third compensation data can be determined in combination with the first compensation data.
[0041] In one possible embodiment, considering that large jumps are prone to occur in featureless areas of the image, and that large jump values are invalid, the mean and standard deviation of several second compensation data corresponding to each region to be scanned in the current scan line can be calculated before determining the third compensation data using the second compensation data. If the difference between a single second compensation data corresponding to the current scan line and the mean is not less than the deviation threshold, the corresponding single second compensation data is discarded, and the new second compensation data corresponding to the current scan line is finally obtained. The deviation threshold is a threshold calculated based on the standard deviation. By using standard deviation calculation, abnormal jump values in the second compensation data can be filtered out, preventing these invalid values from participating in the compensation update, thus ensuring the validity of the second compensation data and improving compensation accuracy.
[0042] Then, using the (new) second compensation data and the first compensation data corresponding to the current scan line, the third compensation data corresponding to the subsequent scan line is determined, and the subsequent scan line is scanned based on the third compensation data. The specific method for determining the third compensation data will be described in detail in subsequent embodiments.
[0043] The first compensation data is essentially a static deformation mapping of the wafer and / or mask in a non-working state, reflecting the global and systematic spatial offset of the wafer and / or mask caused by thermal history or process stress. Before the wafer and / or mask scanning begins, this solution first uses this data to perform coordinate compensation on the scanning trajectory, essentially establishing a relatively accurate initial alignment reference for the entire scanning process. This provides a high-precision starting position for subsequent scanning, avoiding subsequent compensation failure due to excessive initial deviation. During the wafer and / or mask scanning process, the high-speed movement of the motion platform, continuous heat accumulation, and the non-uniformity of local material properties can all induce real-time deformation of the wafer and / or mask. This deformation cannot be fully predicted through static calibration before scanning. The technical solution of this application generates second compensation data by obtaining the coordinate deviation of adjacent dies (patterns) in the current scanning row, essentially using the relative positional deviation between adjacent dies (patterns) as a real-time feedback signal. Because this signal originates from real-time measurements at adjacent positions within the same scan line, it possesses extremely high timeliness and local realism, accurately capturing transient errors caused by temperature drift, vibration, or other random disturbances during the scanning process. During scanning, the first compensation data serves as the basic reference, and the real-time deviation represented by the second compensation data is superimposed to determine the compensation value used for subsequent scan lines (especially those in the same direction), i.e., the third compensation value. This continuous updating of the accuracy of subsequent scan lines enables closed-loop compensation combining the first and second compensation data, significantly improving the overall alignment accuracy of the die (pattern) position. This provides a higher-quality scanning reference for subsequent defect detection (such as bright-field inspection, BFI), effectively suppressing false defect signals or missed defects caused by positioning deviations, ultimately ensuring the accuracy and reliability of wafer and / or mask inspection.
[0044] Example 2
[0045] Based on the above embodiments, this embodiment further explains in detail how to determine the third compensation data corresponding to subsequent scan rows; Figure 6 This is a flowchart of the first compensation data update process. In this embodiment, a serpentine back-and-forth motion is used for line-by-line scanning. Considering the trajectory directions of different scanning operations, the force and motion patterns of the device in the same direction of scanning are basically the same, and the coordinate deviation trends of adjacent lines are also correlated. Therefore, the first compensation data corresponding to the same direction of scanning operations can be updated based on the real-time second compensation data, which can improve the compensation accuracy and avoid errors caused by different motion patterns in opposite directions of scanning.
[0046] Step S21: Determine the next scan line that is in the same direction as the current scan line as the target scan line.
[0047] In this embodiment, when determining the third compensation data using the second compensation data and the first compensation data, the update target is first determined. Specifically, the next scan line in the same direction as the current scan line is determined as the target scan line. For example, assuming the current scan line is the k-th line (scanned from left to right), the next scan line in the same direction as the k-th line is the (k+2)-th line (the (k+1)-th line is scanned in the opposite direction), so the (k+2)-th line is determined as the target scan line.
[0048] Step S22: Update the first compensation data corresponding to the target scan line based on the second compensation data corresponding to the current scan line to obtain the third compensation data corresponding to the target scan line.
[0049] In this embodiment, after the target scan line is determined, the first compensation data corresponding to the target scan line is updated based on the second compensation data corresponding to the current scan line. That is, the second compensation data fed back in real time by the current scan line is superimposed on the first compensation data originally corresponding to the target scan line to obtain the third compensation data corresponding to the target scan line, and the target scan line is scanned based on the third compensation data.
[0050] It is understandable that, since the forces and motion laws of the same-direction scan rows are consistent, the coordinate deviation trends of adjacent same-direction scan rows are highly correlated. Therefore, updating the first compensation data corresponding to the same-direction scan rows based on the real-time second compensation data can improve the compensation accuracy and avoid the error correction caused by the different motion laws of the opposite-direction scans. This makes the residual after compensation converge more and more, that is, the position accuracy of the bare die after compensation becomes higher and higher.
[0051] As can be seen, in this embodiment, considering the trajectory direction of different scanning operations, the force and motion law of the device in the same direction of scanning are basically the same, and the coordinate deviation change trend of adjacent rows is also related. Here, the first compensation data corresponding to the same direction of scanning operation is updated based on the real-time second compensation data, which can improve the compensation accuracy and avoid the error correction caused by the different motion laws of opposite direction scanning.
[0052] Example 3
[0053] Based on the above embodiments, this embodiment combines the following: Figure 7 The flowchart shown illustrates the process of "updating the first compensation data corresponding to the target scan line based on the second compensation data", especially explaining the interval matching relationship between the second compensation data and the first compensation data and the cumulative calculation of the deviation.
[0054] Step S31: Determine the current interval matching relationship between the second interval corresponding to the second compensation data and the first interval of the first compensation data corresponding to the target scan line.
[0055] It should be noted that actual wafers are composed of a large number of independent bare dies, and photomasks are composed of a large number of independent bare die patterns. Manufacturing deviations and equipment processing errors are not uniform in different bare die regions or bare die pattern regions, and the accuracy of the second compensation data and the first compensation data may also be different. Therefore, to ensure scanning accuracy, it is necessary to first establish a matching relationship between the effective range of the second compensation data and the first compensation data on the object to be scanned.
[0056] In this embodiment, the second interval corresponding to the second compensation data includes the region of one or more bare dies corresponding to the wafer and / or the region of one or more bare die patterns corresponding to the mask; the first interval of the first compensation data corresponding to the target scan line is the region of one or more bare dies corresponding to the wafer and / or the region of one or more bare die patterns corresponding to the mask. Specifically, the current interval matching relationship between the second interval corresponding to the second compensation data and the first interval of the first compensation data corresponding to the target scan line is determined, that is, the effective range of the second compensation data on the wafer and / or mask is mapped to the relevant first compensation data interval.
[0057] Step S32: Calculate and sum the deviations of the second compensation data and the first compensation data corresponding to the scan lines in the same direction that have already undergone the scan operation to obtain the total deviation result.
[0058] In this embodiment, deviation calculations are performed on the second compensation data and the first compensation data corresponding to the scan lines in the same direction that have already undergone scanning operations. The corresponding number of deviation calculation results are then accumulated to obtain the total deviation result. For example, assuming that the current scan is the k-th row (scanning from left to right), and the next scan in the same direction is the k+2-th row; the second compensation data (RTA feedback values) at the same position on different bare dies are accumulated to obtain the total deviation result.
[0059] Step S33: Update the first compensation data corresponding to the target scan line using the total deviation result to obtain the third compensation data corresponding to the target scan line.
[0060] In this embodiment, based on the current interval matching relationship determined in step S31, the first compensation data corresponding to the target scan line is updated using the total deviation result obtained in step S32 to obtain the third compensation data corresponding to the target scan line. For example, the total deviation result is added to the first compensation data (premap value of this line) corresponding to the (k+2)th line to obtain the third compensation data (updated premap value) of the (k+2)th line.
[0061] Understandably, for the compensation data of each scan line, the total deviation of the compensation data corresponding to the previous scan lines is calculated by summing, and the first compensation data corresponding to the target scan line is updated using the total deviation. The resulting third compensation data contains the real-time deviation and the cumulative deviation, and the coordinates of the final target scan line will be more accurate. This avoids the situation where the overall scanning positioning error increases as the scan line progresses, and improves the compensation accuracy.
[0062] In this embodiment, for the compensation data of each scan line, the total deviation result of the compensation data corresponding to the previous scan lines is calculated by accumulation, and the first compensation data corresponding to the target scan line is updated using the total deviation result. The resulting third compensation data includes the real-time deviation and the cumulative deviation, and the coordinates of the final target scan line will be more accurate. This avoids the situation where the overall scanning positioning error increases as the scan line progresses, and can improve the compensation accuracy.
[0063] Example 4
[0064] This embodiment is based on the above embodiments, combined with Figure 8 The flowchart shown illustrates the process of generating the second compensation data.
[0065] Step S41: Determine the position data of the area to be scanned in the current scan line obtained by scanning through a preset number of sub-scan bands.
[0066] In this embodiment, a wide full-line scan can be divided into multiple subswaths for parallel acquisition. Specifically, during the scanning process of the scan trajectory, the position data of the area to be scanned in the current scan line is obtained by scanning a preset number of subswaths.
[0067] Step S42: Determine the coordinate deviation between two adjacent areas to be scanned in the current scan row based on the position data.
[0068] In this embodiment, the coordinate deviation between two adjacent areas to be scanned in the current scan row is determined based on the position data obtained in step S41.
[0069] Step S43: Merge the coordinate deviations corresponding to adjacent sub-scan bands, and perform weighted calculations on the corresponding merged data to obtain the second compensation data corresponding to the current scan line.
[0070] In this embodiment, the coordinate deviations corresponding to adjacent sub-scan bands are merged, and the corresponding merged data are weighted to obtain the second compensation data corresponding to the current scan line. In one possible embodiment, the original RTA feedback data has 6 channels (corresponding to 12 sub-scan bands, of which 2 sub-scan bands provide 1 channel of feedback). The FPGA of the synchronization board normalizes the 6 channels of data according to the weight calculation method, that is, synthesizes 1 channel of data as the second compensation data corresponding to the current scan line.
[0071] Understandably, by splitting a wide full-line scan into multiple sub-scan bands for parallel acquisition, the acquisition time for a single scan line can be significantly shortened, thus adapting to the high-speed scanning requirements of large-size wafers and / or photomasks. At the same time, with weighted calculation, reasonable weights can be assigned to different die regions and / or die pattern regions to avoid mutual influence of cross-regional deviations.
[0072] This embodiment divides the full-line scan of a large width into multiple sub-scan bands for parallel acquisition, which greatly shortens the acquisition time of a single scan line. This can adapt to the high-speed scanning requirements of large-size wafers and / or photomasks. At the same time, with the addition of weighted calculation, reasonable weights can be assigned to different die regions and / or die pattern regions to avoid mutual influence of cross-regional deviations.
[0073] Example 5
[0074] The above embodiments describe how real-time compensation data from scanning is used to update the original feedforward compensation data of relevant marker lines during defect detection. This allows for continued scanning using the updated compensation data, improving compensation accuracy and further reducing scanning errors. Furthermore, this embodiment uses wafer scanning as an example to illustrate relevant aspects of wafer defect detection; such as... Figure 9 The diagram illustrates the interaction between the synchronization board and the RTA (Real-Time Analytics) card, which processes real-time compensation data. The FPGA sends the scanned DIE slice information to the RTA card, and the RTA card sends RTA feedback data, i.e., real-time compensation data, to the synchronization board. Figure 10 As shown, this embodiment uses wafer scanning as an example to detail the specific implementation of the defect detection method, including: Step S51: Initialize feedforward compensation data.
[0075] The generation of feedforward compensation data requires scanning the mark line in advance and obtaining the measurement coordinates of the mark point according to the Wafer coordinate system. Simultaneously, based on the theoretical coordinates of the Mark point... Calculate the x-axis offset of each Mark point. ;like Figure 11The diagram shows the x-axis offset. The calculation formula is as follows: .
[0076] Furthermore, based on the position corresponding to the x-axis, linear interpolation is performed on the x-axis (linear extrapolation is used outside the range on both sides of the Mark point, and linear interpolation is used between two Mark points) to determine the initial Premap (feedforward compensation data): ; ; The Premap for the remaining rows expands upwards and downwards using the compensation data obtained from this calculation, maintaining consistency across each column: .
[0077] Step S52: Generate real-time compensation data.
[0078] Then, wafer scanning is performed. Before each line of swath scanning, the synchronization card sends the premap value of the current line to the FPGA, which then compensates for the DIE position. During the swath scanning process, the RTA card sends RTA feedback to the synchronization card in real time. In one possible implementation, the original RTA feedback data has 6 channels (12 subswaths, 2 subswaths as 1 channel). The synchronization card FPGA normalizes the 6 channels of data according to the weight calculation method, that is, it synthesizes them into 1 channel of data.
[0079] It should be noted that real-time compensation data refers to the deviation of a certain position within a DIE relative to the same position within the previous DIE, such as... Figure 12 The figure shows the positional deviation of the marker line at different DIEs.
[0080] Step S53: Calculate the real-time compensation data for the wafer interval corresponding to the feedforward compensation data.
[0081] Furthermore, after each line of swath scanning is completed, the synchronization card reads the synthesized RTA feedback data from the FPGA via DMA (Direct Memory Access). The original RTA feedback data of the k-th line is denoted as... The corresponding x-axis position is denoted as When feedback All are in the feedforward parameter Within the effective range, the corresponding feedforward parameters are obtained by calculating the mean. The RTA feedback is: .
[0082] In one possible implementation, considering that large jumps are likely to occur in featureless areas of the image, and that large jump values are invalid, a standard deviation calculation scheme can be used to process the RTA feedback data and filter out abnormal jump values.
[0083] Step S54: Update the feedforward compensation data.
[0084] The real-time compensation data obtained through the above steps can then be used to update the original feedforward compensation data corresponding to subsequent marker lines, thereby improving compensation accuracy. For example, if the current scan is of row k (scanning from left to right), and the next row in the same direction is row k+2, then the premap value of row k+2 is updated after accumulating the RTA feedback at the same position for different DIEs. Figure 13 As shown, the line scan deviation for different DIEs is updated by accumulation; the specific update formula is as follows: ; in, This represents the index of the j-th premap interval of the r-th DIE in this row.
[0085] Accordingly, starting from line 3, before each swath scan, the updated premap value is sent to the synchronization card FPGA to compensate for the DIE position. Line 3 uses the Premap+RTA value of line 1, line 5 uses the Premap+RTA value of line 3, line 7 uses the Premap+RTA value of line 5, and so on.
[0086] Therefore, during the marker line scanning process, after the synchronization card acquires the RTA feedback data, it first calculates the average value of the RTA feedback values in the premap area corresponding to the marker line of the current row, and then accumulates the RTA feedback at the same position under different DIEs. The accumulated result is added to the premap corresponding to the marker line of the current row to update the premap value corresponding to the next marker line in the same direction. In other words, this application changes the open-loop compensation that relies solely on feedforward compensation data to a closed-loop compensation that relies on both feedforward compensation data and real-time compensation data, which can further reduce the DIE and DIE error during the line scanning process.
[0087] Based on the above embodiments, this embodiment also discloses a scanning device for implementing the scanning method described above.
[0088] Furthermore, this embodiment also discloses a semiconductor device for implementing the scanning method described above.
[0089] In one possible implementation, taking wafer scanning as an example, such as Figure 14As shown, the semiconductor device includes a line scan device, a real-time alignment card (RTA card), and a synchronization board. The line scan device performs a scanning operation on the object to be scanned. The RTA card sends the data obtained by the line scan device through the scanning operation to the synchronization board. The synchronization board executes the scanning method described above based on the data. Specifically, the synchronization board obtains the scanning information corresponding to the current scan line from the line scan device and performs coordinate compensation using pre-prepared feedforward compensation data, i.e., first compensation data. Simultaneously, the synchronization board sends the die position information of the scan information to the RTA card, which generates real-time compensation data, i.e., second compensation data, based on the die's coordinate deviation. The synchronization board then updates the first compensation data corresponding to the next scan line by combining the first and second compensation data, thus obtaining new first compensation data for the next scan line. This combination of feedforward compensation data and dynamic real-time compensation data during the scanning process performs closed-loop compensation for the wafer position, improving the alignment accuracy of the die position.
[0090] Furthermore, this embodiment also discloses a computer-readable storage medium for storing a computer program, which, when executed by a processor, implements the scanning method described above.
[0091] Accordingly, this embodiment also discloses a computer program product, including a computer program / instruction, which, when executed by a processor, implements the scanning method described above.
[0092] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.
[0093] Each of the above modules or units can be implemented through software, hardware, or a combination of both. In this application, "implemented through software" means that the processor reads and executes program instructions stored in memory to implement the functions corresponding to the above modules or units. Here, the processor refers to a processing circuit capable of executing program instructions, including but not limited to at least one of the following: a central processing unit (CPU), a microprocessor, a digital signal processor (DSP), a microcontroller unit (MCU), or an artificial intelligence processor, or other processing circuits capable of running program instructions. In other embodiments, the processor may also include circuits with other processing functions (such as hardware circuits for hardware acceleration, bus and interface circuits, etc.). Processors can be presented as integrated chips, for example, as integrated chips whose processing functions only include executing software instructions, or they can be presented as SoCs (system on a chip), that is, on a single chip, in addition to the processing circuits (usually called "cores") that can run program instructions, there are also other hardware circuits for implementing specific functions (of course, these hardware circuits can also be implemented separately based on ASICs or FPGAs). Correspondingly, in addition to executing software instructions, the processing functions can also include various hardware acceleration functions (such as AI calculations, encoding and decoding, compression and decompression, etc.).
[0094] In this application, "implemented in hardware" means that the functions of the above-mentioned modules or units are implemented through hardware processing circuits that do not have program instruction processing capabilities. These hardware processing circuits can be composed of discrete hardware components or integrated circuits. To reduce power consumption and size, integrated circuits are typically used. Hardware processing circuits can include ASICs (application-specific integrated circuits) or PLDs (programmable logic devices); PLDs can include FPGAs (field-programmable gate arrays), CPLDs (complex programmable logic devices), and so on. These hardware processing circuits can be a single packaged semiconductor chip (e.g., packaged as an ASIC); or they can be integrated with other circuits (e.g., CPUs, DSPs) and packaged into a single semiconductor chip. For example, multiple hardware circuits and a CPU can be formed on a silicon substrate and packaged into a single chip; this type of chip is also called a SoC. Alternatively, circuits for implementing FPGA functions and a CPU can be formed on a silicon substrate and encapsulated into a single chip; this type of chip is also called a SoPC (system on a programmable chip).
[0095] It should be noted that when this application is implemented through software, hardware, or a combination of both, different software or hardware can be used, and it is not limited to using only one type of software or hardware. For example, one module or unit can be implemented using a CPU, while another module or unit can be implemented using a DSP. Similarly, when implemented using hardware, one module or unit can be implemented using an ASIC, while another module or unit can be implemented using an FPGA. Of course, it is not limited to using the same software (e.g., all through a CPU) or the same hardware (e.g., all through an ASIC) to implement some or all modules or units. Furthermore, those skilled in the art will understand that software is generally more flexible but less performant than hardware, while hardware is the opposite. Therefore, those skilled in the art can choose software, hardware, or a combination of both based on actual needs.
[0096] The foregoing preferred embodiments have further illustrated the objectives, technical solutions, and advantages of the present invention. It should be understood that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A scanning method, characterized in that, include: Using the first compensation data corresponding to the object to be scanned, coordinate compensation is performed on the scanning trajectory of the object to be scanned; the object to be scanned is a wafer and / or a mask, the first compensation data is the deviation data between the ideal coordinates and the actual coordinates of the marker lines in the object to be scanned, the marker lines are marker lines set in a preset area of the object to be scanned, and the scanning trajectory includes multiple scanning lines; Generate second compensation data; the second compensation data includes the coordinate deviation of two adjacent areas to be scanned in the current scan line during the scanning process of the scan trajectory, wherein the areas to be scanned are the areas of the bare die corresponding to the wafer and / or the areas of the bare die pattern corresponding to the mask; Using the second compensation data and the first compensation data corresponding to the current scan line, the third compensation data corresponding to the subsequent scan line is determined, and the subsequent scan line is scanned based on the third compensation data.
2. The scanning method according to claim 1, characterized in that, The step of determining the third compensation data corresponding to subsequent scan rows using the second compensation data and the first compensation data corresponding to the current scan row includes: The next scan line that is in the same direction as the current scan line is determined as the target scan line; The first compensation data corresponding to the target scan line is updated based on the second compensation data corresponding to the current scan line to obtain the third compensation data corresponding to the target scan line.
3. The scanning method according to claim 2, characterized in that, The step of updating the first compensation data corresponding to the target scan line based on the second compensation data corresponding to the current scan line to obtain the third compensation data corresponding to the target scan line includes: Determine the current interval matching relationship between the second interval corresponding to the second compensation data and the first interval of the first compensation data corresponding to the target scan line; the second interval includes the region of one or more bare dies corresponding to the wafer and / or the region of one or more bare die patterns corresponding to the mask, and the first interval is the region of one or more bare dies corresponding to the wafer and / or the region of one or more bare die patterns corresponding to the mask; Based on the current interval matching relationship, the first compensation data corresponding to the target scan line is updated using the second compensation data to obtain the third compensation data corresponding to the target scan line.
4. The scanning method according to claim 3, characterized in that, The step of updating the first compensation data corresponding to the target scan line based on the current interval matching relationship using the second compensation data to obtain the third compensation data corresponding to the target scan line includes: For the second compensation data and the first compensation data corresponding to the same direction scan lines that have already undergone scanning operations, the deviation is calculated, and the corresponding number of deviation calculation results are accumulated to obtain the total deviation result; The first compensation data corresponding to the target scan line is updated using the total deviation result to obtain the third compensation data corresponding to the target scan line.
5. The scanning method according to any one of claims 1 to 4, characterized in that, Before determining the third compensation data corresponding to subsequent scan rows using the second compensation data and the first compensation data corresponding to the current scan row, the method further includes: The mean and standard deviation of several second compensation data corresponding to each of the scanned regions in the current scan row are calculated. If the difference between a single second compensation data point corresponding to the current scan line and the mean is not less than the deviation threshold, then the corresponding single second compensation data point is removed, and a new second compensation data point corresponding to the current scan line is finally obtained; the deviation threshold is a threshold calculated based on the standard deviation. The step of determining the third compensation data corresponding to subsequent scan rows using the second compensation data and the first compensation data corresponding to the current scan row includes: Using the new second compensation data and the first compensation data corresponding to the current scan line, the third compensation data corresponding to the subsequent scan line is determined.
6. The scanning method according to claim 1, characterized in that, The process of generating the first compensation data includes: The marking lines of the object to be scanned are scanned, and the deviation data corresponding to the marking lines is obtained based on the deviation between the actual coordinates of the scanned bare die and / or bare die pattern and the ideal coordinates corresponding to the marking lines. By using linear interpolation, the compensation data corresponding to the object to be scanned is calculated using the deviation data, and the first compensation data corresponding to the object to be scanned is obtained.
7. The scanning method according to any one of claims 1 to 3, characterized in that, The process of generating the second compensation data includes: During the scanning process of the scanning trajectory, the position data of the area to be scanned in the current scan line is determined by scanning through a preset number of sub-scanning bands; The coordinate deviation between two adjacent regions to be scanned in the current scan row is determined based on the location data. The coordinate deviations corresponding to adjacent sub-scan bands are merged, and the corresponding merged data are weighted to obtain the second compensation data corresponding to the current scan line.
8. A scanning device, characterized in that, Used to implement the scanning method as described in any one of claims 1 to 7.
9. A semiconductor device, characterized in that, Used to implement the scanning method as described in any one of claims 1 to 7.
10. The semiconductor device according to claim 9, characterized in that, The semiconductor device includes a line scanning device, a real-time alignment card, and a synchronization board; wherein... The line scanning device is used to perform scanning operations on the object to be scanned; The real-time alignment card is used to send the data obtained by the line scanning device through the scanning operation to the synchronization board. The synchronization board is used to perform the scanning method as described in any one of claims 1 to 7 based on the data.
11. A computer-readable storage medium, characterized in that, Used to store a computer program, which, when executed by a processor, implements the scanning method as described in any one of claims 1 to 7.
12. A computer program product, characterized in that, Includes a computer program / instruction that, when executed by a processor, implements the scanning method as described in any one of claims 1 to 7.