A semiconductor testing equipment intelligent calibration method and system

Abstract

The present invention relates to the field of semiconductor detection technology, and specifically to a semiconductor detection equipment intelligent calibration method and system, including: analyzing the relative stability of wafer detection data between different semiconductor device dimension types to obtain local precision control requirements; based on the local precision control requirements, comprehensively analyzing the effective information content of the same wafer detection data in all semiconductor wafers as a whole to obtain the semiconductor precision requirements; analyzing the overall changes in wafer detection data under each semiconductor device dimension type, combining the semiconductor precision requirements of wafer detection data and the signal-to-noise ratio of wafer detection data to obtain the semiconductor control optimization degree; comprehensively considering the semiconductor control optimization degree, adaptively adjusting wafer detection data, and performing intelligent calibration of semiconductor detection equipment. The present invention improves the accuracy of detection results when semiconductor detection equipment detects wafers, making the calibration process of semiconductor detection equipment more intelligent.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor detection technology, and in particular to an intelligent calibration method and system for semiconductor detection equipment. Background Art

[0002] As an important basic material for manufacturing semiconductor products, the quality of wafers will directly affect the quality of corresponding semiconductor products, so wafers need to be quality tested before manufacturing semiconductor products. In order to ensure that the quality test results of wafers using semiconductor testing equipment are more accurate, it is necessary to review the accuracy of semiconductor testing equipment and eliminate errors in a timely manner to achieve calibration of semiconductor testing equipment.

[0003] The prior art usually manually presets various equipment parameters of semiconductor testing equipment when testing wafers to calibrate the semiconductor testing equipment; however, due to certain individual differences between different wafer surfaces and certain correlations between equipment parameters of different items, the accuracy of the test results when the semiconductor testing equipment tests the wafer is reduced, thereby reducing the intelligence of the calibration process of the semiconductor testing equipment. Summary of the invention

[0004] The present invention provides a semiconductor detection equipment intelligent calibration method and system to solve the existing problem: the prior art usually manually presets various equipment parameters of the semiconductor detection equipment when detecting wafers to calibrate the semiconductor detection equipment; and does not consider the individual differences between different wafers and the correlation between different equipment parameters, resulting in the calibration process of the semiconductor detection equipment is not intelligent.

[0005] A semiconductor detection equipment intelligent calibration method and system of the present invention adopts the following technical solutions:

[0006] The present invention proposes a semiconductor detection equipment intelligent calibration method, the method comprising the following steps:

[0007] In each semiconductor device dimension, obtaining a number of wafer inspection data of each semiconductor wafer within a preset time period;

[0008] For a semiconductor wafer, the correlation between wafer inspection data of different semiconductor device dimension types is obtained, the stability of the correlation is analyzed, and the local precision control requirement of each wafer inspection data is obtained; based on the local precision control requirement, the effective information content of the same wafer inspection data in all semiconductor wafers as a whole is comprehensively analyzed to obtain the semiconductor precision requirement of each wafer inspection data;

[0009] Analyze the overall changes in wafer inspection data under each semiconductor device dimension type, and combine the semiconductor accuracy requirements of the wafer inspection data to obtain the semiconductor energy effective consumption of each semiconductor device dimension type; obtain the semiconductor control preference degree of each semiconductor device dimension type based on the semiconductor energy effective consumption and the signal-to-noise ratio of the wafer inspection data under each semiconductor device dimension type;

[0010] The semiconductor control preference degree of each semiconductor device dimension type is comprehensively considered, the wafer inspection data is adaptively adjusted, and the semiconductor inspection equipment is intelligently calibrated.

[0011] Preferably, the method for obtaining the local precision control requirement is:

[0012] For any semiconductor wafer, the semiconductor wafer is combined in all semiconductor device dimension types to obtain a number of wafer dimension data combinations;

[0013] Analyze the correlation relationship representing the wafer surface information in each wafer dimension data combination to obtain the wafer dimension correlation of each wafer dimension data combination;

[0014] According to the stability of the wafer dimension correlation between different wafer dimension data combinations containing the same wafer inspection data, the local accuracy control requirement of each wafer inspection data is obtained.

[0015] Preferably, the method for acquiring the wafer dimension correlation is:

[0016] For any wafer dimension data combination, obtaining a signal-to-noise data sequence of each semiconductor device dimension type in the wafer dimension data combination;

[0017] The similarities between different signal-to-noise data sequences are analyzed to obtain the wafer dimension correlation of the wafer dimension data combination.

[0018] Preferably, the method for obtaining the stability condition is:

[0019] Among all wafer dimension data combinations of the same wafer inspection data, the mean differences in wafer dimension correlations between different wafer dimension data combinations and all wafer dimension data combinations as a whole are compared to obtain stability.

[0020] Preferably, the method for obtaining the semiconductor precision requirement is:

[0021] For any wafer inspection data, obtain the signal-to-noise content variance of the wafer inspection data in all semiconductor wafers;

[0022] According to the signal-to-noise content variance and the local precision control requirement, the semiconductor precision requirement of the wafer inspection data is obtained; the semiconductor precision requirement is negatively correlated with the signal-to-noise content variance.

[0023] Preferably, before calculating the semiconductor precision requirement, the method further includes:

[0024] The local precision control requirements of wafer inspection data in all semiconductor wafers are averaged.

[0025] Preferably, the method for obtaining the semiconductor control preference degree is:

[0026] For any semiconductor device dimension type, the mean signal-to-noise ratio of all semiconductor wafers in the semiconductor device dimension type is obtained; based on the semiconductor energy effective consumption degree and the mean signal-to-noise ratio, the semiconductor control preference degree of the semiconductor device dimension type is obtained.

[0027] Preferably, the method for obtaining the effective consumption of semiconductor energy is:

[0028] For any semiconductor device dimension type, obtaining a wafer inspection data change value of the semiconductor device dimension type in a semiconductor wafer;

[0029] According to the semiconductor precision requirement and wafer inspection data variation value, the semiconductor energy effective consumption of semiconductor equipment dimension type is obtained.

[0030] Preferably, the method for acquiring the wafer inspection data variation value is:

[0031] According to the difference between the inspection data of any two adjacent wafers in the semiconductor wafer according to the semiconductor device dimension type, a local wafer inspection data variation value is obtained.

[0032] The present invention also proposes a semiconductor detection equipment intelligent calibration system, including a memory and a processor, wherein the processor executes a computer program stored in the memory to implement the steps of the above-mentioned semiconductor detection equipment intelligent calibration method.

[0033] The beneficial effects of the technical solution of the present invention are: obtaining the correlation between wafer inspection data of different semiconductor device dimensional types, analyzing the stability of the correlation, and obtaining the local precision control requirement; wherein the local precision control requirement is used to describe the importance of wafer inspection data for adjusting semiconductor inspection equipment, combining the correlation between different device parameters, and reducing the interference of a single device parameter on the calibration process; based on the local precision control requirement, comprehensively analyzing the effective information content of the same wafer inspection data in all semiconductor wafers as a whole, and obtaining the semiconductor precision requirement; wherein the semiconductor precision requirement is used to describe the effective information distribution represented by the wafer inspection data in all semiconductor wafers as a whole, and comprehensively analyzing the effective information content of the same wafer inspection data in all semiconductor wafers as a whole, and obtaining the semiconductor precision requirement. The individual differences between different semiconductor wafers reduce the interference of the individual differences between different semiconductor wafers on the wafer detection data; based on the overall change of wafer detection data under each semiconductor device dimension type and the semiconductor accuracy requirement of wafer detection data, combined with the signal-to-noise ratio of wafer detection data under each semiconductor device dimension type, the semiconductor control preference degree is obtained; wherein the semiconductor control preference degree is used to describe the important situation of the semiconductor device dimension type for adjusting the calibration process of the semiconductor detection equipment, and on the basis of combining the individual differences between different semiconductor wafers, the required adjustment control of each semiconductor device dimension type is analyzed, so that the calibration process of the semiconductor detection equipment is more intelligent. The present invention adaptively adjusts the wafer detection data by comprehensively considering the individual differences between different wafers and the correlation between different equipment parameters, improves the accuracy of the detection results when the semiconductor detection equipment detects the wafer, and makes the calibration process of the semiconductor detection equipment more intelligent. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0035] Figure 1 A flowchart of the steps of a semiconductor detection equipment intelligent calibration method of the present invention;

[0036] Figure 2 A schematic diagram of a semiconductor wafer defect of the present invention;

[0037] Figure 3 The preset radial velocity of the light beam of the present invention is , the wafer rotation speed is Schematic diagram of semiconductor wafer defects;

[0038] Figure 4 The preset radial velocity of the light beam of the present invention is , the wafer rotation speed is Schematic diagram of semiconductor wafer defects. DETAILED DESCRIPTION

[0039] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation, structure, features and effects of a semiconductor detection equipment intelligent calibration method and system proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0040] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0041] The specific scheme of the semiconductor detection equipment intelligent calibration method and system provided by the present invention is described in detail below with reference to the accompanying drawings.

[0042] See also Figure 1 , which shows a flowchart of a method for intelligent calibration of semiconductor detection equipment provided by an embodiment of the present invention, the method comprising the following steps:

[0043] Step S001: in each semiconductor device dimension, a plurality of wafer inspection data of each semiconductor wafer within a preset time period are obtained.

[0044] It should be noted that the prior art usually manually presets various equipment parameters of semiconductor testing equipment when testing wafers to calibrate the semiconductor testing equipment; however, due to certain individual differences between different wafer surfaces and certain correlations between equipment parameters of different items, the accuracy of the test results when the semiconductor testing equipment tests the wafer is reduced, thereby reducing the intelligence of the calibration process of the semiconductor testing equipment.

[0045] It should be further explained that when inspecting semiconductor wafers, the intelligent calibration of semiconductor inspection equipment is completed by adjusting the two equipment parameters of the radial scanning speed of the laser equipment on the wafer surface and the rotation speed of the wafer surface. Therefore, this embodiment uses these two equipment parameters as an example of semiconductor equipment dimension type for analysis, where these two equipment parameters can be determined according to the specific implementation situation. Please refer to Figure 2 , which shows a schematic diagram of semiconductor wafer defects, Figure 2The blank squares in the middle are defective areas, the gray squares are normal areas, the minimum circumscribed circle is the schematic outline of the wafer edge, and the dotted x and y axes are the two-dimensional coordinate system constructed through the center of the wafer; please refer to Figure 3 , which shows that the preset beam radial velocity is , the wafer rotation speed is Schematic diagram of semiconductor wafer defects, Figure 3 The blank squares in the middle are defective areas, the gray squares are normal areas, and the smallest circumscribed circle is the schematic outline of the wafer edge; please refer to Figure 4 , which shows that the preset beam radial velocity is , the wafer rotation speed is Schematic diagram of semiconductor wafer defects, Figure 4 The blank squares in the middle are defective areas, the gray squares are normal areas, and the smallest circumscribed circle is the schematic outline of the wafer edge.

[0046] In a specific implementation of an embodiment of the present invention, a method for acquiring wafer inspection data is as follows: the speed of radial scanning of the laser device on the wafer surface is recorded as the beam radial speed; and the rotation speed of the wafer surface is recorded as the wafer rotation speed. In the device parameter database of the semiconductor inspection device, several speed data and signal-to-noise ratio data of several semiconductor wafers in two semiconductor device dimension types of beam radial speed and wafer rotation speed are obtained in the past month, and each speed data is recorded as wafer inspection data. Each wafer inspection data corresponds to a signal-to-noise ratio data.

[0047] It is particularly noted that this embodiment takes the example of recording wafer inspection data of each semiconductor wafer once every 1 minute for a total of 1 hour to obtain a number of wafer inspection data; in addition, the quantity and recording time of wafer inspection data can be determined according to specific implementation conditions.

[0048] So far, a number of wafer inspection data of a number of semiconductor wafers in a number of semiconductor device dimension types are obtained through the above method.

[0049] Step S002: For a semiconductor wafer, obtain the correlation between wafer inspection data of different semiconductor device dimension types, analyze the stability of the correlation, and obtain the local precision control requirement of each wafer inspection data; based on the local precision control requirement, comprehensively analyze the effective information content of the same wafer inspection data in all semiconductor wafers as a whole, and obtain the semiconductor precision requirement of each wafer inspection data.

[0050] It should be noted that, for any semiconductor wafer, the wafer inspection data under different semiconductor equipment dimension types represent the equipment parameters of different items of the semiconductor inspection equipment, and these equipment parameters will directly affect the clarity of the wafer surface image information expression; and under normal circumstances, there are certain differences in the impact of these equipment parameters on the clarity of the wafer surface image, which causes the corresponding wafer surface clarity content to fluctuate with different amplitudes, thereby producing similar content to different degrees; therefore, for a semiconductor wafer, the correlation of wafer inspection data between different semiconductor equipment dimension types can be obtained, the stability of the correlation can be analyzed, and the local precision control requirement of each wafer inspection data can be obtained; wherein, if the local precision control requirement is greater, it means that the wafer inspection data makes the image information expression of the corresponding wafer surface clearer and more stable, reflecting that the wafer inspection data is more important for adjusting semiconductor inspection equipment.

[0051] It should be further explained that, although the local precision control requirement reflects the adaptability of wafer inspection data to semiconductor inspection equipment to a certain extent, it can only reflect the matching of wafer inspection data to a single semiconductor wafer; and the differences between different semiconductor wafers themselves will cause certain differences in the corresponding more adaptable wafer inspection data; therefore, based on the local precision control requirement, the effective information content of the same wafer inspection data in all semiconductor wafers as a whole can be comprehensively analyzed to obtain the semiconductor precision requirement of each wafer inspection data; among them, the greater the semiconductor precision requirement, the more effective information distribution represented by the wafer inspection data in all semiconductor wafers as a whole, the lower the interference of individual differences between different semiconductor wafers on the wafer inspection data, reflecting that the wafer inspection data is closer to the ideal value and the lower the degree of adjustment.

[0052] Preferably, in some implementations of the embodiments of the present invention, the method for obtaining the local precision control requirement is: for any semiconductor wafer, combine the semiconductor wafer in all semiconductor device dimension types to obtain a number of wafer dimension data combinations; analyze the correlation relationship that characterizes the wafer surface information in each wafer dimension data combination to obtain the wafer dimension correlation of each wafer dimension data combination; obtain the local precision control requirement of each wafer detection data based on the stability of the wafer dimension correlation between different wafer dimension data combinations containing the same wafer detection data. The specific process is:

[0053] 1. Obtain wafer dimension data combination.

[0054] All wafer inspection data of semiconductor wafers between different semiconductor equipment dimension types are arbitrarily combined in pairs to obtain several wafer dimension data combinations.

[0055] It should be noted that each wafer dimension data combination includes wafer inspection data of two semiconductor device dimension types.

[0056] 2. Obtain wafer dimension correlation.

[0057] It should be noted that the wafer dimension data combination represents the data combination between different wafer detection data of different semiconductor device dimension types; since the combinations of different wafer detection data have different degrees of clarity in expressing the wafer surface information, the correlation between different wafer dimension data combinations that represent the wafer surface information can be analyzed to obtain the wafer dimension correlation of each wafer dimension data combination; among them, the greater the wafer dimension correlation, the more clearly the numerical combination between different wafer detection data in the wafer dimension data combination can express the wafer surface information.

[0058] Preferably, in some implementations of the embodiments of the present invention, the method for obtaining the wafer dimension correlation is: for any wafer dimension data combination, obtain the signal-noise data sequence of each semiconductor device dimension type in the wafer dimension data combination; analyze the similarities between different signal-noise data sequences to obtain the wafer dimension correlation of the wafer dimension data combination. The specific process is as follows:

[0059] Any wafer detection data in the wafer dimension data combination is recorded as the target wafer detection data, and all wafer dimension data combinations containing the target wafer detection data are used as the same type wafer data combinations of the target wafer detection data; among all the same type wafer data combinations of the target wafer detection data, the sequence after the signal-to-noise ratio data of other wafer detection data except the target wafer detection data are arranged in descending order is used as the signal-to-noise data sequence of the target wafer detection data; the signal-to-noise data sequence of each wafer detection data in the wafer dimension data combination is obtained; the Pearson correlation coefficient between the signal-to-noise data sequences of all wafer detection data in the wafer dimension data combination is used as the wafer dimension correlation of the wafer dimension data combination; and the wafer dimension correlation of each wafer dimension data combination is obtained.

[0060] It should be noted that, if the wafer dimension correlation is greater, it means that the numerical combination between different wafer detection data in the wafer dimension data combination can more clearly express the surface information of the wafer.

[0061] It is particularly noted that the acquisition of the Pearson correlation coefficient is a well-known technique and will not be described in detail in this embodiment.

[0062] 3. Obtain local precision control requirements.

[0063] As an example, the local precision control requirement of the target wafer dimension data can be calculated by the following formula:

[0064]

[0065] In the formula, Indicates the local accuracy control requirement of target wafer dimension data; The number of all similar wafer data combinations representing target wafer dimension data; The first one represents the target wafer dimension data The wafer dimension correlation of the same type of wafer data combination; The mean value of the wafer dimension correlation of all similar wafer data combinations representing the target wafer dimension data; Indicates taking the absolute value; represents the linear normalization function.

[0066] It should be noted that in the formula It indicates the stability of the wafer dimension correlation between different wafer dimension data combinations containing the same wafer inspection data; if the local precision control requirement is greater, it means that the target wafer inspection data makes the image information expression of the corresponding wafer surface less clear and stable, reflecting that the target wafer inspection data is more important for adjusting the semiconductor inspection equipment.

[0067] Preferably, the method for obtaining the semiconductor precision requirement is: for any wafer inspection data, obtain the signal-to-noise content variance of the wafer inspection data in all semiconductor wafers; obtain the semiconductor precision requirement of the wafer inspection data according to the signal-to-noise content variance and the local precision control requirement; the semiconductor precision requirement is negatively correlated with the signal-to-noise content variance. The specific process is as follows:

[0068] The variance of all signal-to-noise ratio data of the target wafer inspection data in all semiconductor wafers is taken as the signal-to-noise content variance; the mean of the local precision control requirements of the target wafer inspection data in all semiconductor wafers is taken as the mean of the wafer comprehensive local precision control requirement; the normalized value of the ratio between the mean of the wafer comprehensive local precision control requirement and the signal-to-noise content variance is taken as the semiconductor precision requirement of the target wafer inspection data.

[0069] It should be noted that the greater the demand for semiconductor precision, the more effective information distribution represented by the wafer inspection data in all aspects of the semiconductor wafer as a whole, and the lower the interference of individual differences between different semiconductor wafers on the wafer inspection data, which means that the wafer inspection data is closer to the ideal value and the lower the degree of adjustment.

[0070] It is particularly noted that this embodiment uses The linear normalization function performs normalization.

[0071] At this point, the semiconductor accuracy requirement of each wafer inspection data is obtained through the above method.

[0072] Step S003: Analyze the overall changes in wafer inspection data under each semiconductor device dimension type, and combine the semiconductor accuracy requirements of the wafer inspection data to obtain the effective semiconductor energy consumption of each semiconductor device dimension type; based on the effective semiconductor energy consumption and the signal-to-noise ratio of the wafer inspection data under each semiconductor device dimension type, obtain the semiconductor control preference degree of each semiconductor device dimension type.

[0073] It should be noted that in the process of adjusting the parameters of each semiconductor wafer for a period of time, due to the differences in the material distribution of the semiconductor wafer itself, there are certain differences in the amount of energy consumed by the semiconductor detection equipment when adjusting the equipment parameters of different semiconductor wafers; therefore, the overall changes in the wafer detection data under each semiconductor equipment dimension type are analyzed, and the semiconductor accuracy requirements of the wafer detection data are combined to obtain the effective semiconductor energy consumption of each semiconductor equipment dimension type; the greater the effective semiconductor energy consumption, the more energy the semiconductor equipment dimension type consumes for the entire semiconductor wafer, the greater the efficiency of the semiconductor equipment dimension type in adjusting the equipment parameters by detecting the wafer, and the less likely the energy consumed by the semiconductor equipment dimension type is to be wasted.

[0074] It should be further explained that, in addition to the individual differences between different semiconductor wafers, each semiconductor wafer itself will undergo parameter adjustments for a period of time to ensure relatively accurate detection results; and the parameter adjustments during this period will also continuously affect the situation in which the content on the wafer surface can be clearly expressed; therefore, the semiconductor control preference of each semiconductor device dimension type can be obtained based on the effective semiconductor energy consumption and the signal-to-noise ratio of the wafer detection data under each semiconductor device dimension type; if the semiconductor control preference is greater, it means that when the semiconductor device dimension type detects the wafer, it is more necessary to control the device parameters under the semiconductor device type, reflecting that the semiconductor device dimension type is more important for the calibration process of the semiconductor detection equipment.

[0075] Preferably, in some implementations of the embodiments of the present invention, the method for obtaining the effective consumption of semiconductor energy is: for any semiconductor device dimension type, obtaining the wafer inspection data variation value of the semiconductor device dimension type in the semiconductor wafer; according to the semiconductor precision requirement and the wafer inspection data variation value, obtaining the effective consumption of semiconductor energy of the semiconductor device dimension type. The specific process is:

[0076] Preferably, in some implementations of the embodiments of the present invention, the method for obtaining the wafer inspection data variation value is: according to the difference between any two adjacent wafer inspection data in the semiconductor wafer according to the semiconductor device dimension type, the local wafer inspection data variation value is obtained. The specific process is as follows:

[0077] The first The absolute value of the difference between any two adjacent wafer inspection data of the semiconductor device dimension type in the semiconductor wafer is used as the local wafer inspection data change value. As an example, the following formula can be used to calculate the first Semiconductor energy efficiency of each semiconductor device dimension type:

[0078]

[0079] In the formula, Indicates The semiconductor energy efficiency consumption of each semiconductor device type; Represents the number of all semiconductor wafers; Indicates The semiconductor device dimension type is in The average semiconductor accuracy requirement of all wafer inspection data in a semiconductor wafer; Indicates The semiconductor device dimension type is in The number of change values ​​of all local wafer inspection data in a semiconductor wafer; Indicates The semiconductor device dimension type is in Semiconductor wafer The local wafer inspection data change value.

[0080] It should be noted that Indicates The semiconductor device dimension type is in The larger the wafer inspection data variation value is, the smaller the wafer inspection data variation value is. The semiconductor device dimension type is in The more energy is consumed in each semiconductor wafer, the greater the effective consumption of semiconductor energy is. The more energy a semiconductor device type consumes for the entire semiconductor wafer, the The greater the efficiency of adjusting device parameters by inspecting wafers, the greater the efficiency of the semiconductor device dimension type. The smaller the type of semiconductor equipment, the less likely it is to waste the energy it consumes.

[0081] Preferably, in some implementations of the embodiments of the present invention, the method for obtaining the semiconductor control preference degree is: for any semiconductor device dimension type, obtaining the mean signal-to-noise ratio of all semiconductor wafers in the semiconductor device dimension type; and obtaining the semiconductor control preference degree of the semiconductor device dimension type according to the semiconductor energy effective consumption degree and the mean signal-to-noise ratio. The specific process is:

[0082] The normalized value of the product of the mean signal-to-noise ratio and the semiconductor energy efficient consumption is used as the semiconductor control preference level of the semiconductor device dimension type.

[0083] It should be noted that the greater the degree of semiconductor control preference, the more it is necessary to control the device parameters under the semiconductor device type when the semiconductor device dimension type is used to detect the wafer, which reflects that the semiconductor device dimension type is more important for the calibration process of the semiconductor detection equipment.

[0084] At this point, the semiconductor control preference level for each semiconductor device dimension type is obtained through the above method.

[0085] Step S004: Comprehensively consider the semiconductor control preference level of each semiconductor device dimensional type, adaptively adjust the wafer inspection data, and perform intelligent calibration of the semiconductor inspection equipment.

[0086] It should be noted that the semiconductor control preference degree reflects the degree to which the corresponding wafer inspection data of each semiconductor device dimensional type needs to be adjusted during the calibration process of adjusting the semiconductor inspection equipment by inspecting the wafer, thereby affecting the final calibration accuracy; therefore, the semiconductor control preference degree of each semiconductor device dimensional type can be comprehensively considered, the wafer inspection data can be adaptively adjusted, and the semiconductor inspection equipment can be intelligently calibrated.

[0087] In a specific implementation of the embodiment of the present invention, the specific process of intelligent calibration is: taking any semiconductor device dimension type as an example, a semiconductor control preference threshold is preset. , if the semiconductor device dimension type semiconductor control preference is greater than , preset the wafer inspection data of semiconductor device dimension type as , and The product of the semiconductor control preference degree of the semiconductor device dimension type is used as the adaptive wafer inspection data of the semiconductor device dimension type.

[0088] It is particularly noted that in this embodiment Take this as an example, It may depend on the specific implementation situation.

[0089] It should be noted that the process of obtaining adaptive wafer inspection data is the intelligent calibration process of semiconductor inspection equipment.

[0090] Through the above steps, a semiconductor detection equipment intelligent calibration method is completed.

[0091] Another embodiment of the present invention provides an intelligent calibration system for semiconductor detection equipment, the system comprising a memory and a processor, and when the processor executes the computer program stored in the memory, the above method steps S001 to S004 are performed.

[0092] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A semiconductor testing equipment intelligent calibration method, characterized in that: The method comprises the following steps: In each semiconductor device dimension, obtaining a number of wafer inspection data of each semiconductor wafer within a preset time period; For a semiconductor wafer, the correlation between wafer inspection data of different semiconductor device dimension types is obtained, and the stability of the correlation is analyzed to obtain the local precision control requirement of each wafer inspection data; based on the local precision control requirement, the effective information content of the same wafer inspection data in all semiconductor wafers as a whole is comprehensively analyzed to obtain the semiconductor precision requirement of each wafer inspection data, including: For any wafer inspection data, the variance of all signal-to-noise ratio data of the wafer inspection data in all semiconductor wafers is taken as the signal-to-noise content variance; the mean of the local precision control requirements of the wafer inspection data in all semiconductor wafers is taken as the mean of the wafer comprehensive local precision control requirements; the normalized value of the ratio between the mean of the wafer comprehensive local precision control requirements and the signal-to-noise content variance is taken as the semiconductor precision requirement of the wafer inspection data; Analyze the overall changes in wafer inspection data under each semiconductor equipment dimension type, and combine the semiconductor accuracy requirements of wafer inspection data to obtain the semiconductor energy effective consumption of each semiconductor equipment dimension type, including: The first The absolute value of the difference between any two adjacent wafer inspection data of the semiconductor device dimension type in the semiconductor wafer is used as the local wafer inspection data change value to calculate the The semiconductor energy efficiency of each semiconductor device dimension type includes: In the formula, Indicates The semiconductor energy efficiency consumption of each semiconductor device type; Represents the number of all semiconductor wafers; Indicates The semiconductor device dimension type is in The average semiconductor accuracy requirement of all wafer inspection data in a semiconductor wafer; Indicates The semiconductor device dimension type is in The number of change values ​​of all local wafer inspection data in a semiconductor wafer; Indicates The semiconductor device dimension type is in Semiconductor wafer The local wafer inspection data change value; According to the semiconductor energy effective consumption degree and the signal-to-noise ratio of the wafer inspection data under each semiconductor device dimension type, the semiconductor control preference degree of each semiconductor device dimension type is obtained; The semiconductor control preference degree of each semiconductor device dimension type is comprehensively considered, the wafer inspection data is adaptively adjusted, and the semiconductor inspection equipment is intelligently calibrated.

2. According to claim 1, a semiconductor detection equipment intelligent calibration method is characterized in that: The method for obtaining the local precision control requirement is: For any semiconductor wafer, the semiconductor wafer is combined in all semiconductor device dimension types to obtain a number of wafer dimension data combinations; Analyze the correlation relationship representing the wafer surface information in each wafer dimension data combination to obtain the wafer dimension correlation of each wafer dimension data combination; According to the stability of the wafer dimension correlation between different wafer dimension data combinations containing the same wafer inspection data, the local accuracy control requirement of each wafer inspection data is obtained.

3. The semiconductor detection equipment intelligent calibration method according to claim 2, characterized in that: The method for obtaining the wafer dimension correlation is: For any wafer dimension data combination, obtaining a signal-to-noise data sequence of each semiconductor device dimension type in the wafer dimension data combination; The similarities between different signal-to-noise data sequences are analyzed to obtain the wafer dimension correlation of the wafer dimension data combination.

4. The semiconductor detection equipment intelligent calibration method according to claim 2, characterized in that: The method for obtaining the stability condition is: Among all wafer dimension data combinations of the same wafer inspection data, the mean differences in wafer dimension correlations between different wafer dimension data combinations and all wafer dimension data combinations as a whole are compared to obtain stability.

5. The semiconductor detection equipment intelligent calibration method according to claim 1, characterized in that: The method for obtaining the semiconductor control preference degree is: For any semiconductor device dimension type, the mean signal-to-noise ratio of all semiconductor wafers in the semiconductor device dimension type is obtained; based on the semiconductor energy effective consumption degree and the mean signal-to-noise ratio, the semiconductor control preference degree of the semiconductor device dimension type is obtained.

6. A semiconductor testing equipment intelligent calibration system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the computer program is executed by a processor, the steps of a semiconductor detection equipment intelligent calibration method as described in any one of claims 1 to 5 are implemented.

Images