Disclosure of Invention
The present invention is directed to solving, at least to some extent, one of the technical problems in the related art. Therefore, an objective of the present invention is to provide a method for rapidly detecting flatness variation within a set range of a wafer.
The invention provides a method for detecting flatness change in a wafer setting range. According to an embodiment of the invention, the method comprises: acquiring flatness data of at least one line on a wafer, wherein the flatness data comprises at least one of thickness data and shape data; determining a radial data curve of each line according to the flatness data; calculating each radial data curve through a Douglas-Puck algorithm to obtain a characteristic line data curve corresponding to each line; calculating the slope of a straight line determined by two adjacent points on the data curve of each characteristic line and taking an absolute value to obtain an absolute value of the slope; and comparing the maximum value or the average value of the absolute value of the slope within the set range on the wafer with a detection slope threshold value to obtain a detection result. The method can quickly detect the flatness change in a set range, particularly the flatness change in a small range, and the flatness change is used as a quality judgment basis for subsequent processing of the wafer, so that yield and cost loss caused by invalid processing procedures can be avoided, and the problem that the yield of a client is influenced due to the fact that a specific defect wafer is abnormal at the client so as to influence the satisfaction degree of a client is avoided.
According to an embodiment of the invention, the wafer is a wafer that has not been subjected to final sorting during processing.
According to an embodiment of the present invention, the wafer is a wafer after a wire cutting, grinding or polishing step in the processing process.
According to an embodiment of the invention, the line satisfies at least one of the following conditions: the flatness data of each of the lines includes flatness data of a plurality of points; the line is a straight line passing through the center of the wafer; the number of the lines is 4-8; the included angle between any two adjacent lines is equal.
According to an embodiment of the present invention, the maximum distance between a connection line between two adjacent points on the radial data curves and a curve between the two adjacent points is defined as a chord distance, and the douglas-pock algorithm threshold of each radial data curve is determined by the following steps: and taking the chord distances of all adjacent two points on one radial data curve and calculating the average value of the chord distances, wherein the threshold value of the Douglas-Puck algorithm of the radial data curve is n times of the average value of the chord distances, and n is 2, 3, 4 or 5.
According to an embodiment of the present invention, the setting range is a circular ring area concentric with the wafer or a circular area concentric with the wafer.
According to an embodiment of the present invention, the detection slope threshold is determined by: respectively obtaining a plurality of detection slope absolute values of the wafer in a detection step and a plurality of final sorting slope absolute values in a final sorting step, wherein the detection slope absolute values and the final sorting slope absolute values are in one-to-one correspondence; determining the detection slope threshold based on a plurality of the detection slope absolute values, a plurality of the final sorting slope absolute values, and the final sorting slope threshold in the final sorting step.
According to an embodiment of the present invention, determining the detection slope threshold based on a plurality of the detection slope absolute values, a plurality of the final sorting slope absolute values, and the final sorting slope threshold in the final sorting step includes: establishing a coordinate system by taking the final sorting slope absolute value as an abscissa and the detection slope absolute value as an ordinate, obtaining a plurality of coordinate points based on the plurality of detection slope absolute values and the plurality of final sorting slope absolute values, and performing linear fitting on the plurality of coordinate points to obtain a fitting straight line; and substituting the final sorting slope threshold value into the fitting straight line to calculate to obtain the detection slope threshold value.
According to an embodiment of the present invention, the obtaining the detection result in the method includes: if the maximum value or the average value of the absolute values of the slopes is larger than the threshold value, judging that the wafer is unqualified, and converting the wafer into a negative film; and if the maximum value or the average value of the absolute values of the slopes is less than or equal to the threshold, judging that the wafer is qualified, and continuing to perform subsequent processing.
Detailed Description
The following describes embodiments of the present invention in detail. The following examples are illustrative only and are not to be construed as limiting the invention. The examples, where specific techniques or conditions are not indicated, are to be construed according to the techniques or conditions described in the literature in the art or according to the product specifications.
It should be noted that "final sorting" in the present invention refers to a step of performing final detection after all processes have been completed on a wafer, and the thickness and shape of a silicon wafer are scanned and detected over the entire surface in the final sorting step, and the precision can reach tens of nanometers.
The invention provides a method for detecting flatness change in a wafer setting range. According to an embodiment of the invention, the method comprises the steps of:
s1: flatness data of at least one line on a wafer is acquired, and the flatness data comprises at least one of thickness data and shape data.
In this step, the flatness data of the wafer may be detected by using conventional equipment and methods, for example, the method includes but is not limited to a capacitive or double-sided laser measurement method, and scanning measurement is performed on lines on the wafer, specifically, flatness data of a plurality of points may be tested on each line, and the flatness data of the plurality of points may be plotted as a curve, so that corresponding calculation may be performed on the detected data. In some embodiments, the test equipment may output raw data of thicknesses and shapes of tens or hundreds of points, and the detection data of each line may be plotted to obtain a corresponding data curve.
According to the embodiments of the present invention, the position of the test line may be selected according to actual detection requirements, and in some specific embodiments, the line is a straight line passing through the center of the wafer. Therefore, the wafer is linearly scanned to obtain test data, the operation is simple and convenient, the flatness change of the wafer can be better reflected by the test result, and the test precision is higher. According to an embodiment of the present invention, the number of the lines may be 4 to 8, specifically, 4, 6, 8, etc. Therefore, excessive measurement and data calculation are not needed, the flatness change of the wafer can be well reflected by the test result, and the test accuracy is high.
According to the embodiment of the invention, the distribution of the test lines can be selected according to actual detection requirements, in some specific embodiments, a proper line position can be selected according to a set range, generally, the lines are uniformly distributed in the set range as much as possible, and specifically, the included angles between any two adjacent lines are equal. Therefore, the detection result can better reflect the flatness change condition in the set range, and the accuracy is higher. In one embodiment, referring to fig. 1, the range is set to be the whole wafer, the number of the test lines is 8, and the included angle between two adjacent test lines is 22.5 degrees. Therefore, lines are uniformly distributed, and the testing accuracy is high.
It is understood that the main processing steps of the wafer are cutting, cleaning, edge grinding, alkali etching, polishing, chemical mechanical polishing, and final sorting of the silicon wafer, and according to some embodiments of the present invention, the wafer used in the detection method of the present invention is a wafer that is not subjected to final sorting during the processing, and specifically may be a wafer after the wire cutting, grinding or polishing step during the processing. Therefore, the flatness change of the wafer can be effectively detected in the initial processing stage, the wafer which does not accord with the use standard is removed as soon as possible, the processing steps are saved, the cost waste is avoided, and meanwhile, the processing yield can be improved.
S2: and determining a radial data curve of each line according to the flatness data.
In this step, the raw data of each line output by the test equipment is converted into radial data of each line. Specifically, the radial data of each line uses the distance from each point on the line to the center of the wafer as the position coordinate of the point. In some embodiments, the line passes through the center of the wafer, and the center of the wafer divides the line into two parts, which may be 0 point, and the distance between a point on one part of the line divided by the center of the wafer and the center of the wafer is positive and the distance between a point on the other part and the center of the wafer is negative.
The specific form of the radial data is described in detail below by taking a line as an example passing through the center of a wafer, and referring to fig. 1 specifically, by taking a line 1 as an example, the center of the wafer is taken as a point 0, the distance from a point on a radius from the center of the wafer to the direction indicated by the arrow takes a positive value, and the distance from a point on a radius opposite to the direction of the arrow to the center of the wafer takes a negative value, specifically, if the distances from the point a and the point B on the line 1 to the center of the wafer are both 50mm, then when the distance is expressed in the form of the radial data, the position coordinate of the point a is 50mm, and the position coordinate of the point B is-50 mm.
Furthermore, the position of the point on each line is used as an abscissa, the flatness data of the point on each line is used as an ordinate, and a curve is drawn, so that a radial data curve of each line can be obtained.
S3: and calculating each radial data curve through a Douglas-Puck algorithm to obtain a characteristic line data curve corresponding to each line.
Specifically, the douglas-pock algorithm is an algorithm that approximately represents a curve as a series of points and reduces the number of points. The specific calculation process is as follows: (1) connecting a straight line segment between the head point and the tail point of the curve, wherein the straight line segment is a chord of the curve; (2) obtaining a point C with the maximum distance from the straight line segment on the curve, and calculating the distance between the point C and the straight line segment; (3) comparing the distance with a preset threshold value of the Douglas-Puck algorithm, and if the distance is smaller than the threshold value of the Douglas-Puck algorithm, taking the straight line segment as the approximation of a curve, and finishing the processing of the straight line segment; (4) if the distance is larger than the threshold value of the Douglas-Puck algorithm, dividing the curve into two sections by using a point C, and respectively carrying out the processing of the steps (1) to (3) on the two sections of the curve; (5) when all the curves are processed, the broken lines formed by all the dividing points are connected in sequence, and the broken lines can be used as the approximation of the curves.
Specifically, referring to fig. 2, a straight line segment ab is connected between two points a and b at the beginning and the end of the curve to obtain a point c with the maximum distance from the straight line segment on the curve, the distance L between the point c and the straight line segment ab is greater than a preset douglas-pock algorithm threshold, the curve is divided into two segments ac and bc by using the point c, then a straight line segment ac is connected between the two points a and c at the beginning and the end of the curve, a straight line segment bc is connected between the two points b and c at the beginning and the end of the curve to obtain a point d with the maximum distance from the straight line segment ac on the curve and a point e with the maximum distance from the straight line segment bc on the curve, the distance L1 between the point d and the straight line segment ac is less than the preset douglas-pock algorithm threshold, the straight line segment ac is used as an approximation of the curve, the distance L2 between the point e and the straight line segment bc is greater than the preset douglas-pock algorithm threshold, and then respectively obtaining a point f with the maximum distance from the straight line section ec on the curve and a point g with the maximum distance from the straight line section eb on the curve, wherein the distance between the point f and the straight line section ec is smaller than a preset Douglas-Pock algorithm threshold, the straight line section ec is used as the approximation of the curve, the distance between the point g and the straight line section eb is larger than the preset Douglas-Pock algorithm threshold, the curve is divided into two sections eg and gb by the point g, and the eg and the gb are both straight line sections, and the curve is processed.
According to an embodiment of the present invention, the threshold in the douglas-pock algorithm (i.e., the douglas-pock algorithm threshold) may be selected according to actual situations. In some embodiments, the maximum distance from a connection line between two adjacent points on the radial data curves to a curve between the two adjacent points is defined as a chord distance, and the threshold of the douglas-pock algorithm of each radial data curve is determined by: and taking the chord distances of all adjacent two points on one radial data curve and calculating the average value of the chord distances, wherein the threshold value of the Douglas-Puck algorithm of the radial data curve is n times of the average value of the chord distances, and n is 2, 3, 4 or 5.
S4: and calculating the slope of a straight line determined by two adjacent points on the data curve of each characteristic line and taking the absolute value to obtain the absolute value of the slope.
Specifically, in this step, a straight line may be determined by two adjacent points on each feature line data curve, and an equation of the straight line passing through the two points may be calculated according to coordinates of the two points in a coordinate system in which the position of the point on each feature line data curve is used as an abscissa and flatness data of the point on each feature line data curve is used as an ordinate, so that an absolute value of a slope may be determined.
S5: and comparing the maximum value or the average value of the absolute value of the slope within the set range on the wafer with a detection slope threshold value to obtain a detection result.
According to the embodiment of the invention, the specific setting range can be set according to the detection requirement. Specifically, the set range may be a circular ring region concentric with the wafer or a circular region concentric with the wafer. In some embodiments, the predetermined range may be a ring having a radial dimension of 10mm,15mm,20mm,25mm,30mm or other values from the edge of the wafer, the predetermined range may be a circle having a radial dimension of 10mm,15mm,20mm,25mm,30mm,40mm,50mm,60mm or other values from the center of the wafer, and the predetermined range may be a whole wafer.
Note that the "absolute value of the slope in the set range" herein refers to the absolute value of the slope of a straight line defined by two adjacent points located in the set range.
According to the embodiment of the invention, the detection slope threshold value can be calculated by collecting a series of data or set according to actual operation experience. In some embodiments, a series of data may be collected in advance, and then the detection slope threshold may be calculated according to the collected data.
Specifically, the detection slope threshold may be determined by the following steps: respectively obtaining a plurality of detection slope absolute values of the wafer in a detection step and a plurality of final sorting slope absolute values in a final sorting step, wherein the detection slope absolute values and the final sorting slope absolute values are in one-to-one correspondence; determining the detection slope threshold based on a plurality of the detection slope absolute values, a plurality of the final sorting slope absolute values, and the final sorting slope threshold in the final sorting step.
It should be noted that the detection slope absolute values and the final sorting slope absolute values may be detection slope absolute values and final sorting slope absolute values corresponding to a plurality of wafers, or detection slope absolute values and final sorting slope absolute values corresponding to a plurality of points on one wafer, and the one-to-one correspondence between the detection slope absolute values and the final sorting slope absolute values means that the detection slope absolute values and the final sorting slope absolute values are respectively obtained corresponding to one wafer or one point, and if the detection slope absolute values respectively correspond to wafers numbered 1 to 10, the final sorting slope absolute values are also wafers corresponding to numbers 1 to 10.
According to an embodiment of the present invention, determining the detection slope threshold based on a plurality of the detection slope absolute values, a plurality of the final sorting slope absolute values, and the final sorting slope threshold in the final sorting step may include: establishing a coordinate system by taking the final sorting slope absolute value as an abscissa and the detection slope absolute value as an ordinate, obtaining a plurality of coordinate points based on the plurality of detection slope absolute values and the plurality of final sorting slope absolute values, and performing linear fitting on the plurality of coordinate points to obtain a fitting straight line; and substituting the final sorting slope threshold value into the fitting straight line to calculate to obtain the detection slope threshold value.
In an embodiment, referring to fig. 3, the abscissa is the absolute value of the final sorting slope, the ordinate is the absolute value of the detection slope, the abscissa of each coordinate point is the absolute value of the final sorting slope and the absolute value of the detection slope at the same point on one wafer or one wafer, a plurality of coordinate points are obtained through detection and calculation, and linear fitting is performed on the plurality of coordinate points, so as to obtain a fitting straight line, where the formula of the fitting straight line shown in fig. 3 is y 0079x +0.2918, R2R0.7857, as can be seen2The decision coefficient is close to 1, which indicates that the fitting degree of the final sorting slope absolute value and the detection slope absolute value is high, and further, when the final sorting slope threshold is 200, the final sorting slope threshold is substituted into the above fitting straight line formula for calculation, namely, x is 200, the formula y is 0079x +0.2918, and y is 1.8718, namely, the detection slope threshold is 1.8718.
According to an embodiment of the present invention, the obtaining of the detection result in the method may include: if the maximum value or the average value of the absolute value of the slope is larger than the detection slope threshold, determining that the wafer is unqualified, and converting the wafer into a negative film; and if the maximum value or the average value of the absolute values of the slopes is less than or equal to the detection slope threshold, judging that the wafer is qualified, and continuing to perform subsequent processing. Therefore, the quality of the wafer can be better controlled, unqualified wafers can be removed in the initial machining process, the yield and cost loss caused by invalid machining procedures can be avoided, the abnormal condition of the specific defect silicon wafer at the client side can be avoided, and the client satisfaction can be effectively improved.
The method can quickly detect the flatness change in a set range, particularly the flatness change in a small range, and is used as a quality judgment basis for the subsequent processing of the silicon wafer, so that the yield and cost loss caused by invalid processing procedures can be avoided, and the influence on the customer satisfaction caused by the yield of a specific defect wafer occurring in a customer end abnormally is avoided.
The following describes the method for detecting flatness variation within a wafer setting range in detail by taking an embodiment as an example, and the specific steps are as follows: loading shape data output by detection equipment, determining a radial data curve according to the shape data output by the detection equipment, then calculating a characteristic line data curve through a Douglas-Pock algorithm (Ramer-Douglas-Peucker algorithm), calculating absolute values of slopes of two adjacent points on each characteristic line data curve, comparing the absolute values of the slopes in a set range of all characteristic lines, taking a maximum value or an average value, and detecting the magnitude of a slope threshold value, if the maximum value or the average value of the slope absolute values is greater than the detection slope threshold value, judging that the wafer is unqualified, and converting the wafer into a negative film; and if the maximum value or the average value of the absolute values of the slopes is less than or equal to the detection slope threshold, judging that the wafer is qualified, and continuing to perform subsequent processing.
Specifically, fig. 4 shows a radial shape data curve a1 and a corresponding feature line data curve B1 (only feature points obtained by the douglas-pock algorithm are shown in the figure, and a broken line obtained by connecting two adjacent feature points by a straight line segment is shown in the figure as a feature line data curve) and an absolute slope value C1 of a polished wafer in an embodiment, and fig. 5 shows a 3D graph (left figure) of 8-bar shape data output by the wafer inspection apparatus and a 3D graph (right figure) of an absolute slope value obtained by the method of the present invention.
Fig. 6 shows a radial data curve a2 of a shape of a polished wafer and a corresponding data curve B2 of a feature line (only feature points obtained by the douglas-pock algorithm are shown in the figure, and a broken line connecting two adjacent feature points by a straight line segment is shown in the figure as a data curve of a feature line) and an absolute value of a slope C2 in another embodiment, and fig. 7 shows a 3D graph (left figure) of 8 bar-shaped data output by the wafer inspection apparatus and a 3D graph (right figure) of an absolute value of a detected slope obtained by the method of the present invention.
Comparing fig. 7 and fig. 5, after the douglas-pock algorithm and the calculation of the absolute value of the slope, the difference in the shape of the local flatness can be clearly recognized.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.