JP2015211133A - Cmp analyzer, cmp analysis method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a CMP analyzer for supporting identification of a cause of generation of an abnormal wafer.SOLUTION: A CMP analyzer 10 comprises: a data acquisition unit 11 that is connected to a CMP device 20 and acquires, from the CMP device 20, data having a plurality of variables on an operation situation and data having a plurality of variables on a polishing profile of a wafer; an arithmetic control unit 12 that obtains, from the variables on the operation situation, a variable whose degree of correlation with an objective variable selected from the plurality of variables on the polishing profile is equal to or higher than a predetermined threshold, and takes the obtained variable as an explanation variable for the objective variable; and an output unit 14 that outputs the explanation variable obtained by the arithmetic control unit 12. The arithmetic control units 12 obtains the explanation variable on the basis of data at the time of polishing generating a normal wafer and data at the time of polishing generating an abnormal wafer.

Description

  The present invention relates to an apparatus for analyzing the operating status of a CMP apparatus and the polishing status of a wafer.

  Conventionally, it has been known to acquire data such as operation status from a CMP (Chemical Mechanical Polishing) apparatus that polishes a semiconductor wafer and analyze the data to analyze the operation status of the CMP apparatus.

  Data acquired from the CMP apparatus includes several hundreds of data even in the polishing process alone, and high skill is required for data analysis. There is no established method for troubleshooting when trouble occurs in the CMP apparatus. Accordingly, the actual situation is that the way of viewing the data differs depending on the experience value of each individual, and the priority of the data to be checked (for example, whether it is caused by hardware or caused by consumables) is different.

  Patent Document 1 discloses an invention of an apparatus engineering data feature extraction method for outputting a representative value of apparatus engineering data (EES data) having a strong correlation with product quality fluctuations.

JP 2012-79755 A

  In the invention described in Patent Document 1, it is possible to output device engineering data having a strong correlation with quality fluctuations, but in order to analyze the output device engineering data and perform troubleshooting, a certain skill is still required. Is done.

  In view of the above background, an object of the present invention is to provide a CMP analyzer that supports the identification of the cause of abnormal wafer generation.

  The CMP analysis apparatus of the present invention is connected to a CMP apparatus, a data acquisition unit that acquires data of a plurality of variables related to an operating state and data of a plurality of variables related to a polishing profile of a wafer from the CMP apparatus, and the polishing profile Control for obtaining a variable whose correlation level with a target variable selected from a plurality of variables is greater than or equal to a predetermined threshold value from among the variables related to the driving situation, and using the obtained variable as an explanatory variable of the target variable And an output unit that outputs the explanatory variable obtained by the calculation control unit, the calculation control unit is configured to generate data when a normal wafer is generated by polishing and when an abnormal wafer is generated. Based on this, the explanatory variable is obtained. Here, the objective variable may be at least one of a polishing rate, in-plane uniformity, and continuous stability.

  According to the present invention, an explanatory variable having a high correlation with an objective variable selected from a plurality of variables related to the polishing profile is obtained based on data when a normal wafer and an abnormal wafer are generated. The explanatory variable obtained in this way is highly likely to indicate the cause when the abnormal wafer is generated, and it is easy to investigate the cause of the abnormal wafer generated by grasping this explanatory variable. be able to. The data relating to the driving situation includes EES data, data indicating the usage time of consumables, and the like.

  In the CMP analyzer of the present invention, the arithmetic control unit determines whether an abnormality has occurred in the CMP apparatus based on the data of the objective variable acquired by the data acquisition unit, and no abnormality has occurred. Sometimes the data of the explanatory variable is stored in the storage means, and when an abnormality occurs in the CMP apparatus, the data of the explanatory variable acquired when the abnormality occurs is compared with the data of the explanatory variable stored in the storage means. An explanatory variable in which a difference larger than the threshold value is obtained may be obtained, and the output unit may output the explanatory variable obtained by the arithmetic control unit.

  In this way, by comparing the data of the explanatory variables acquired and stored in the normal state with the data of the explanatory variables at the time of occurrence of the abnormality, the explanatory variable that is considered to indicate the cause of the abnormality from among a plurality of explanatory variables Can be narrowed down.

  The CMP analysis apparatus of the present invention includes an objective variable setting unit that sets a desired value of the objective variable, and the arithmetic control unit obtains a function of the objective variable using the explanatory variable as a parameter, and based on the function Then, the value of the explanatory variable for setting the objective variable to the desired value may be obtained, and the output unit may output the value of the explanatory variable corresponding to the desired value.

  In this way, by outputting the value of the explanatory variable corresponding to the desired value of the objective variable, it is possible to propose the setting of the CMP apparatus for obtaining the desired polishing profile, and to support the development of the process.

  In the CMP analysis apparatus of the present invention, the calculation control unit explains, in the data relating to the operation state, data whose change amount when a normal wafer is generated and when an abnormal wafer is generated is a predetermined threshold value or more. You may narrow down as a variable candidate.

  Data with a small amount of change when a normal wafer is generated and when an abnormal wafer is generated can be interpreted as data not related to the quality of the wafer. According to the present invention, by excluding data with a small change amount, candidates for explanatory variables can be narrowed down and the calculation load for obtaining the explanatory variables can be reduced.

  The CMP apparatus of the present invention has a configuration including the CMP analysis apparatus. By providing the CMP analysis apparatus, it is possible to provide a CMP apparatus that supports investigation of the cause when an abnormal wafer is generated and development of a new process.

  The CMP analysis method of the present invention is a method for analyzing the operation status of CMP by a CMP analysis device connected to the CMP device, wherein the CMP analysis device receives variable data related to the operation status from the CMP device, and A step of acquiring data of a plurality of variables relating to a polishing profile of a wafer, and a variable in which the CMP analyzer has a correlation level higher than a predetermined threshold value with respect to a target variable selected from the plurality of variables relating to the polishing profile Is obtained from the variables related to the operation status, and the obtained variable is used as the explanatory variable of the objective variable, and the data when the normal wafer is generated by polishing and when the abnormal wafer is generated Obtaining the explanatory variable based on the output, and the CMP analyzer outputs the explanatory variable obtained by the arithmetic control unit That and a step.

  The program according to the present invention is a program for analyzing a CMP operation state by a CMP analysis device connected to a CMP apparatus, and the program includes data on variables related to the operation state from the CMP apparatus, and wafer data. A step of obtaining data of a plurality of variables related to a polishing profile, and a variable whose correlation with a target variable selected from the plurality of variables related to the polishing profile is a predetermined threshold or more And obtaining the explanatory variable based on data when a normal wafer is generated by polishing and when an abnormal wafer is generated by using the obtained variable as an explanatory variable of the objective variable. And a step of outputting the explanatory variable obtained by the arithmetic control unit.

  The CMP analysis apparatus of the present invention has an effect that it is possible to support the investigation of a cause when an abnormal wafer is generated and the development of a new process.

It is a figure which shows the structure of the CMP analyzer of this Embodiment. It is a figure which shows the relationship between the data acquired from the CMP apparatus, the objective variable, and the explanatory variable. It is a figure which shows the example of CMP apparatus. It is a figure which shows the operation | movement which calculates | requires explanatory variables with CMP analyzer. It is a figure which shows the operation | movement which shows the cause of abnormality of CMP apparatus by CMP analyzer. It is a figure which shows the operation | movement of the process development support by a CMP analyzer. It is a figure which shows an example of a polishing rate.

  The CMP analyzer according to the embodiment of the present invention will be described below.

(First embodiment)
FIG. 1 is a diagram showing a configuration of a CMP analysis apparatus 10 according to an embodiment of the present invention. The CMP analysis apparatus 10 is connected to the CMP apparatus 20 and acquires data on variables related to operating conditions and data on variables related to polishing profiles from the CMP apparatus 20. Here, the data acquired from the CMP apparatus 20 includes data when a normal wafer is generated and data when an abnormal wafer is generated. Further, information for identifying whether data is generated when a normal wafer is generated or data when an abnormal wafer is generated may be included.

  The CMP analysis apparatus 10 includes a data acquisition unit 11, a calculation control unit 12, an input unit 13, an output unit 14, and a storage unit 15. The data acquisition unit 11 has a function of acquiring data related to the operation status and data related to the polishing profile from the CMP apparatus 20. The data acquisition unit 11 stores the acquired data in the storage unit 15.

  Predetermined data among the data relating to the polishing profile is selected as an objective variable for determining whether or not the wafer is normal. The objective variable may be set by the user or automatically set by the CMP analysis apparatus 10. When the CMP analysis apparatus 10 sets the objective variable, for example, the user sets whether the wafer is normal or abnormal, and the CMP analysis apparatus 10 sets a variable whose value is significantly different between the normal wafer and the abnormal wafer. It can be set as the structure selected as a variable.

  FIG. 2 is a diagram showing the relationship between the data acquired from the CMP apparatus 20, the objective variable, and the explanatory variable. As shown in FIG. 2, the objective variable is selected from the variables related to the polishing profile. In the present embodiment, three variables of polishing rate, in-plane uniformity, and continuous stability are selected as objective variables.

  The arithmetic control unit 12 has a function of obtaining, as an explanatory variable, a variable having a high correlation with the target variable from among variables related to the driving situation. Although there are several hundred or more variables related to the driving situation, the arithmetic control unit 12 obtains, as an explanatory variable, a variable having a high correlation with the target variable from among these many variables. As shown in FIG. 2, for each of the three objective variables, several highly correlated variables are obtained as explanatory variables.

  The arithmetic control unit 12 has a function of obtaining an explanatory variable that is considered to be the cause of the abnormality when an abnormality occurs in the CMP apparatus 20. The processing performed by the arithmetic control unit 12 will be described in the description of the operation of the CMP analyzer 10.

  The input unit 13 has a function of receiving input of various types of information (for example, setting of the objective variable described above) from the user. The output unit 14 has a function of outputting the explanatory variable obtained by the arithmetic control unit 12 and information on the cause of the abnormality to the user.

  The hardware configuration of the CMP analyzer 10 is a computer including a CPU, a RAM, a ROM, a hard disk, and the like. By causing a computer to execute a program for CMP analysis, the CMP analysis apparatus 10 having each function shown in FIG. 1 is realized. Such a program is also included in the scope of the present invention.

  FIG. 3 is a diagram showing a configuration of the CMP apparatus 20 connected to the CMP analysis apparatus 10 in the present embodiment. The CMP apparatus 20 shown in FIG. 3 is an example, and the CMP analysis apparatus 10 can be applied to apparatuses other than the CMP apparatus 20 shown in FIG.

  The CMP apparatus 20 is a polishing apparatus that polishes a semiconductor wafer used in a semiconductor device. The CMP apparatus 20 includes a polishing head, a polishing head drive motor, a polishing pad, a turntable, a table drive motor, a dresser, a dresser drive motor, a slurry supply device, and the like.

  The CMP apparatus 20 includes variables relating to operating conditions, for example, torque current D1 when rotating the table drive motor, pad wear amount D2 of the polishing pad, torque current D3 for dresser drive, and slurry (abrasive) flow rate D4. There are a CAP pressure (Center Airbag Pressure) D5 supplied to the polishing head, a torque current D6 of the polishing head, a rotational speed D7 of the turntable, and the like. Although FIG. 3 shows an example, the data actually obtained from the CMP apparatus 20 relates to hundreds of variables or more. As data of the variables related to the polishing profile obtained from the CMP apparatus 20, there are data of the above-described polishing rate D8, in-plane uniformity D9, and continuous stability D10.

  FIG. 4 is a diagram illustrating the operation of the CMP analysis apparatus 10 of the present embodiment. First, the CMP analysis apparatus 10 defines normal / abnormal wafers (S10). For example, the wafer is defined as abnormal when the polishing rate at a certain point on the wafer is higher than a normal threshold or a predetermined threshold value immediately after replacement of consumables. This definition may be performed by the user, or may be automatically performed by the CMP analyzer 10 based on normal / abnormal data of the wafer.

  Next, the CMP analyzer 10 sets an objective variable (S11). In the present embodiment, three target variables are set: polishing rate, in-plane uniformity, and continuous stability. As described above, the objective variable may be set by the user or the CMP analysis apparatus 10.

  The CMP analysis apparatus 10 acquires data relating to the operation status and polishing profile data from the CMP apparatus 20 (S12), and stores the acquired data in the storage unit 15. The arithmetic control unit 12 of the CMP analyzer 10 uses the data stored in the storage unit 15 to obtain a variable (explanatory variable) having a high correlation with the objective variable. Specifically, first, the variables related to the driving situation are narrowed down (S13). Specifically, if the amount of change between the data when a normal wafer is generated and the data when an abnormal wafer is generated is smaller than a predetermined threshold, the data is excluded from the explanatory variable candidates. The data whose change amount is equal to or greater than a predetermined threshold is set as a candidate for explanatory variables.

  Subsequently, the arithmetic control unit 12 of the CMP analysis apparatus 10 obtains an explanatory variable by performing a correlation analysis between the objective variable and the explanatory variable candidate (S14). When performing correlation analysis, processing is performed using the average value of the variables as the origin so that the presence or absence of correlation does not change depending on how the origin is determined. The arithmetic control unit 12 calculates a correlation coefficient between the objective variable and the explanatory variable candidates using the rank correlation coefficient, and extracts a predetermined number of variables as explanatory variables in descending order of the correlation coefficient. The rank correlation coefficient is used because it is not affected by outliers and because the data is not always normally distributed.

  In the present embodiment, an example has been described in which the arithmetic control unit 12 automatically obtains explanatory variables in descending order of correlation coefficient. For example, the correlation is greater than or equal to a predetermined threshold (for example, 0.8). It is good also as a structure which shows a number to a user in order with a high correlation coefficient, and makes a user select an explanatory variable.

  Subsequently, the CMP analysis apparatus 10 stores the explanatory variable obtained by the calculation control unit 12 in the storage unit 15 (S15). Note that the CMP analysis apparatus 10 may notify the user of an explanatory variable having a high correlation with the target variable by outputting the explanatory variable from the output unit 14.

  FIG. 5 is a diagram illustrating an operation in which the CMP analysis apparatus 10 presents the cause when an abnormality occurs in the CMP apparatus 20. During the operation of the CMP apparatus 20, the CMP analysis apparatus 10 acquires data relating to the operation status and data relating to the polishing profile from the CMP apparatus 20 (S20). The CMP analyzer 10 analyzes data on the polishing profile and determines whether or not an abnormality has occurred (S21).

  When it is determined that there is no abnormality (NO in S21), the CMP analyzer 10 stores the data acquired from the CMP apparatus 20 in the storage unit 15 as normal data (S22).

  When it is determined that an abnormality has occurred in the CMP apparatus 20 (YES in S21), the CMP analyzer 10 reads from the storage unit 15 normal data of the explanatory variable corresponding to the objective variable determined to be abnormal. The read normal data is compared with the explanatory variable data when the abnormality occurs. Then, among the plurality of explanatory variables corresponding to the objective variable, an explanatory variable having a significant difference from the normal data is extracted. A t-test can be used for testing a significant difference between normal data and abnormal data. That is, the t-test is first performed between normal data, and a “significance level that does not show a difference” is set, and then the normal data and abnormal data are compared to exceed the significance level. An explanatory variable having a significant difference is extracted (S23).

  Then, the CMP analyzer 10 outputs an explanatory variable showing a significant difference from the normal data as the cause of the abnormality (S24). At this time, when presenting a plurality of explanatory variables, they may be presented in order from the explanatory variables having the largest difference from the normal data, or the explanatory variables are presented together with the difference data from the normal data. Also good. As a result, the user can quickly investigate the cause when an abnormality occurs in the CMP apparatus 20.

  The configuration and operation of the CMP analyzer 10 according to the present embodiment have been described above. The CMP analyzer 10 uses the data when a normal wafer is generated and when an abnormal wafer is generated to obtain an explanatory variable having a high correlation with the objective variable. Explanatory variables related to can be determined appropriately.

  Further, the normal data of the obtained explanatory variable is stored, and when an abnormality occurs in the CMP apparatus 20, the data of the explanatory variable obtained at that time is compared with the normal data to determine the abnormality. Explanatory variables that are likely to be the cause can be presented, which can be useful for troubleshooting when an abnormality occurs.

(Second Embodiment)
Next, the CMP analyzer 10 of the second embodiment will be described. The basic configuration of the CMP analyzer 10 of the second embodiment is the same as that of the CMP analyzer 10 of the first embodiment, and the correlation of the objective variable is similar to that of the first embodiment. Find high explanatory variables. The CMP analysis apparatus 10 according to the second embodiment supports development of a new process using explanatory variables.

  FIG. 6 is a diagram illustrating an operation of the CMP analysis apparatus 10 according to the second embodiment. The CMP analysis apparatus 10 of the second embodiment first receives an input of a desired value of the objective variable (S20). The desired value of the objective variable is, for example, a value related to a target polishing profile when a new polishing profile is developed.

  Subsequently, the arithmetic control unit 12 of the CMP analyzer 10 calculates the value of the explanatory variable corresponding to the desired value of the objective variable (S21). Here, a method for calculating the value of the explanatory variable corresponding to the desired value of the objective variable will be described.

FIG. 7 is a graph showing an example of the polishing rate, in which the vertical axis indicates the polishing rate and the horizontal axis indicates the position on the wafer by the distance from the substrate center. The polishing rate RR (X0) at a certain point (X = X0) can be approximated by the following function F.
RR (X0) = F (X0, X1, X2, X3,...) (1)
Here, the variable X can be considered, for example, X0 = X coordinate, X1 = pad depletion amount, X2 = center airbag pressure (CAP) input value, X3 = turntable rotation speed,... Etc. As an example of the function F, for example,
F = a × X0 + b × X1 + c × X2 + d × X3 + (2)
A first order polynomial can be considered.

  The arithmetic control unit 12 obtains the coefficients a to d in the equation (2) from the data of the objective variable and the explanatory variable. The arithmetic control unit 12 obtains variables X0, X1,... For achieving a desired value of the input objective variable. At this time, X0 = X coordinate and X1 = pad wear amount are variables that cannot be changed by the user side, so they are fixed, and X2 = center airbag input value, X3 = turntable rotation speed, etc. Find the optimum value for it.

  The CMP analyzer 10 outputs the optimum value of the explanatory variable obtained by the arithmetic control unit 12 (S22) and presents it to the user. Thus, the user can obtain a clue as to which setting of the CMP apparatus 20 should be changed when developing a new polishing profile by referring to the provided explanatory variables. Processes can be developed efficiently.

  It should be noted that the result of the polishing performed by the CMP apparatus 20 using the above-described optimum value obtained by the CMP analysis apparatus 10 is compared with the result of the simulation of the CMP analysis apparatus 10, and the parameter of the above formula (2) is obtained. May be updated to improve the accuracy of the model.

  Further, the CMP analyzer 10 of the second embodiment may perform a process of presenting the cause of an abnormality when an abnormality occurs, as in the first embodiment.

  The configuration and operation of the CMP analyzer 10 according to the embodiment of the present invention have been described with reference to the embodiment. However, the present invention is not limited to the above-described embodiment.

  In the above-described embodiment, the CMP analysis apparatus 10 separated from the CMP apparatus 20 has been described as an example. However, the CMP analysis apparatus 10 may be included in the CMP apparatus 20.

  Although not mentioned in the above embodiment, the user may be able to set the calculation area when performing the CMP analysis. For example, the measurement area may be set for each measurement point, or may be set for each predetermined area where a plurality of measurement points are collected.

  The present invention has an effect that it is possible to appropriately obtain an explanatory variable that is highly likely to cause an abnormal wafer, and is useful as an apparatus for analyzing the operating state of a CMP apparatus.

DESCRIPTION OF SYMBOLS 10 CMP analyzer 11 Data acquisition part 12 Operation control part 13 Input part 14 Output part 15 Storage part 20 CMP apparatus

Claims (8)

A data acquisition unit connected to a CMP apparatus and acquiring data of a plurality of variables related to an operation state and data of a plurality of variables related to a polishing profile of a wafer from the CMP apparatus;
A variable whose degree of correlation with a target variable selected from a plurality of variables related to the polishing profile is equal to or higher than a predetermined threshold value is obtained from the variables related to the driving situation, and the obtained variable is an explanatory variable of the target variable. An arithmetic control unit to
An output unit for outputting the explanatory variable obtained by the arithmetic control unit;
With
The calculation control unit is a CMP analyzer that obtains the explanatory variable based on data when a normal wafer is generated by polishing and when an abnormal wafer is generated. The arithmetic control unit determines whether or not an abnormality has occurred in the CMP apparatus based on the objective variable data acquired by the data acquisition unit, and stores the explanatory variable data when no abnormality has occurred. When the abnormality occurs in the CMP apparatus, the data of the explanatory variable acquired when the abnormality occurs is compared with the data of the explanatory variable stored in the storage means, and a difference larger than a predetermined threshold value is generated. Find the explanatory variable
The CMP analysis apparatus according to claim 1, wherein the output unit outputs an explanatory variable obtained by the arithmetic control unit. An objective variable setting unit for setting a desired value of the objective variable;
The calculation control unit obtains a function of the objective variable having the explanatory variable as a parameter, and obtains a value of the explanatory variable for setting the objective variable to the desired value based on the function,
The CMP analysis apparatus according to claim 1, wherein the output unit outputs a value of an explanatory variable corresponding to the desired value.   The calculation control unit narrows down data having a predetermined amount or more as a candidate for the explanatory variable when the normal wafer is generated and when the abnormal wafer is generated, among the data relating to the operation state. 4. The CMP analyzer according to any one of items 1 to 3.   The CMP analysis apparatus according to claim 1, wherein the objective variable is at least one of a polishing rate, in-plane uniformity, and continuous stability.   A CMP apparatus comprising the CMP analysis apparatus according to claim 1. A method for analyzing a CMP operation status by a CMP analyzer connected to a CMP apparatus,
The CMP analyzer acquires data of a plurality of variables related to an operating state and data of a plurality of variables related to a polishing profile of a wafer from the CMP apparatus;
The CMP analyzer obtains a variable whose correlation with an objective variable selected from a plurality of variables related to the polishing profile is a predetermined threshold value or more from the variables related to the driving situation, and the obtained variable An explanatory variable of the objective variable, the step of obtaining the explanatory variable based on data when a normal wafer is generated by polishing and when an abnormal wafer is generated;
The CMP analyzer outputting the explanatory variables obtained by the arithmetic control unit;
A CMP analysis method comprising: A program for analyzing a CMP operation state by a CMP analyzer connected to a CMP apparatus, the program comprising:
Obtaining a plurality of variable data relating to the operation status from the CMP apparatus and a plurality of variable data relating to the polishing profile of the wafer;
A variable whose degree of correlation with a target variable selected from a plurality of variables related to the polishing profile is equal to or higher than a predetermined threshold value is obtained from the variables related to the driving situation, and the obtained variable is an explanatory variable of the target variable. Obtaining the explanatory variable based on data when a normal wafer is generated by polishing and when an abnormal wafer is generated; and
Outputting an explanatory variable obtained by the arithmetic control unit;
A program that executes.