JPH09191032A - Process abnormality monitoring method and device - Google Patents

Abstract

(57) 【Abstract】 PROBLEM TO BE SOLVED: To automatically determine abnormality in a semiconductor wafer process and determine whether review is necessary. SOLUTION: Defect / foreign particle coordinate data is stored in a data storage means DB from a visual inspection apparatus GK and a foreign material inspection apparatus PK via a network Nt, and XY coordinate system data is rθ.
The calculation unit 9 calculates the defect density in each divided element by converting it into the divided elements of the coordinate system, and determines the distribution abnormality by the chi-square distribution. In the case of abnormal distribution, the calculation unit 9 calculates the average value and standard deviation of the distribution grouping and the defect density distribution in each group, compares these values with a preset conditional expression, and the main control unit 12 performs a review. It is determined whether or not there is a need, and when it is determined that a review is necessary, an alarm is issued by the alarm generator 10, the defect coordinate data is automatically transferred to the review station, and the standard deviation and the average value are stored in the data storage means DB.
To be stored.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a process abnormality monitoring method and apparatus, and is particularly effective when applied to automatic monitoring of appearance, defects detected by a foreign substance inspection device, and abnormal distribution of foreign substances in a semiconductor device manufacturing process. Technology.

[0002]

2. Description of the Related Art In a semiconductor manufacturing process, it is essential to find a cause of a defect early and feed it back to a process and a manufacturing apparatus in order to maintain and improve the yield. For this purpose, it is important to detect defects by the inspection device and analyze the inspection data.

According to a study made by the inventor of the present invention, for example, a defect such as a short circuit or a disconnection generated in a wafer process or a foreign substance is detected by a visual inspection device to which image processing is applied or a dark field illumination by a laser beam. It is automatically inspected by the inspection device.

Then, these inspection devices output coordinate data in the semiconductor wafer of the defect and foreign substance candidates to determine what kind of defect the defect and foreign substance candidate is actually.
That is, the classification of disconnection, short circuit, foreign matter, etc. is performed by a person by moving the stage again to the output coordinate position and observing an enlarged image by a metal microscope or a scanning electron microscope.

Hereinafter, this work will be referred to as a review. Usually, at this time, a predetermined classification number is given as data to each defect and foreign substance candidate.

Then, a classification number based on a review is added to the coordinate data and transferred to an analysis system via a network to determine whether the number of defects is within a reference value and whether there is an abnormal bias in the distribution of defects. There is.

The final defect determination is performed on each semiconductor chip by electrical testing after the completion of the entire wafer process, but an abnormal deviation (hereinafter referred to as cluster) from the connection statistics of defective chips. The failure cause is estimated by deciding whether or not these are indicated, grouping these, and determining the similarity of the processes performed on the semiconductor wafers of each group and the similarity with the past data of the defective chip map. .

As an example of a patent that describes the above-mentioned analysis system in detail, Japanese Patent Laid-Open No. 3-4405.
Japanese Patent Laid-Open No. 6-61314 is an example of a patent that describes in detail the method for estimating the cause of a defect from a defective chip map.

[0009]

However, the present inventor has found that the abnormality determination method by monitoring the process as described above has the following problems.

That is, since the above-described review work is a manual work, it takes time, which is a shackle for improving the throughput of the inspection process.

Further, even if an attempt is made to increase the processing capacity by increasing the number of reviewing devices, accurate classification requires knowledge of semiconductor manufacturing processes and structures, and a wide range of insight regarding the causal relationship between defects and defects. Since the work is entrusted to a specific person with specialized knowledge, there is a problem that there are not enough review experts.

Further, when the review is not performed steadily but only when there is an abnormality such as when the number of defects exceeds a predetermined reference value, even if the number of defects is the same, the way of distribution is Especially, since it is often closely related to the process defect, there is a possibility that the information on the number of defects is insufficient for the abnormality detection.

The object of the present invention is to quantify any abnormalities in the appearance, defects output from a foreign matter inspection device, and foreign matter data, and to automatically determine whether there is an abnormality in the distribution and the need for review. It is to provide a process abnormality monitoring method and apparatus capable of making a positive determination.

The above and other objects and novel features of the present invention will be apparent from the description of this specification and the accompanying drawings.

[0015]

SUMMARY OF THE INVENTION Among the inventions disclosed in the present application, the outline of a representative one will be briefly described.
It is as follows.

That is, the process abnormality monitoring method of the present invention comprises the steps of measuring the distribution state of defects and particles on the semiconductor wafer, and converting the measured distribution state into at least one one-dimensional data. Dividing the one-dimensional data into arbitrary dividing elements,
The step of calculating the defect foreign matter density, which is the density of defects and foreign matter in the dividing element, and the defect foreign matter density distribution, which is the distribution of defects and foreign matter density, is abnormal based on the calculated defective foreign matter density in the dividing element. Determining whether or not there is one, grouping the one-dimensional data,
The step of calculating the defect particle density for each grouped group, the step of calculating the average value and the standard deviation of the defect number and the particle number for each grouped group, and the calculated average value and standard deviation From the semiconductor wafer to determine whether or not visual confirmation work for defects or foreign matter is required.

Thus, it is possible to automatically monitor the appearance, the defects output from the foreign matter inspection device, and the abnormalities in the distribution of foreign matter data, and automatically determine whether or not a review is necessary.

The process abnormality monitoring method of the present invention is
The one-dimensional data is position data in the circumferential direction, the radial direction, the horizontal direction, and the vertical direction on the semiconductor wafer.

With this, it is possible to easily detect the characteristics of the uneven distribution of defects and foreign matter in the single-wafer processed semiconductor wafer.

Further, in the process abnormality monitoring method of the present invention, the grouping step includes a step of smoothing a defect foreign matter density distribution, a step of calculating a first derivative of the defective foreign matter density, and a step of calculating the defective foreign matter density. And the step of obtaining the characteristic points of the distribution of the first derivative.

Further, the process abnormality monitoring method of the present invention is
The characteristic point of the distribution of the first derivative of the defective foreign matter density is that the value of the first derivative of the defective foreign matter density changes from negative to positive.

Further, in the process abnormality monitoring method of the present invention, the step of determining whether or not there is an abnormality in the defective foreign matter density distribution is the square of the difference between the defective foreign matter density and the average value of the defective foreign matter density. Value divided by the average value of foreign matter density and chi 2
It consists of the step of comparing with the percentile of the power distribution.

The process abnormality monitoring method of the present invention is
The step of determining whether or not there is an abnormality in the defect foreign matter density distribution comprises a step of comparing a contrast value obtained from the maximum value and the minimum value of the defect foreign matter density with a reference value of a preset contrast value. Is.

Further, in the process abnormality monitoring method of the present invention, the step of judging whether or not the visual confirmation work of defects or foreign matters on the semiconductor wafer is necessary, the average value and the standard deviation and the preset threshold value are set. Comprising the steps of comparing.

The process abnormality monitoring method of the present invention is
The step of determining whether or not the visual confirmation work of defects or foreign matters on the semiconductor wafer is required includes the step of comparing an average value of the number of defects and foreign matters in the entire semiconductor wafer with a preset threshold value. Is.

As a result, it is possible to quantify the defects and foreign substances in each group on the semiconductor wafer, and it is possible to more reliably perform the automatic determination as to whether or not a review is necessary.

Further, the process abnormality monitoring apparatus of the present invention includes a data storage means for storing measurement data of distribution states of defects and foreign substances on a semiconductor wafer, and at least one one-dimensional distribution state stored in the data storage means. First calculating means for converting into data, dividing means for dividing the one-dimensional data converted by the first calculating means into arbitrary dividing elements, defects in the dividing elements divided by the dividing means, and Based on the defective foreign matter density calculated by the second computing means for calculating the defective foreign matter density, which is the density of the foreign matter, the defect and the defective foreign matter density distribution, which is the distribution of the foreign matter density, are abnormal. First control means for determining whether or not there is 1 converted by the first calculation means
Third computing means for grouping the dimensional data, and the third computing means
Fourth computing means for calculating the defect foreign matter density for each group grouped by the computing means, and the average value and the standard of the number of defects and the foreign matter number for each group grouped by the third computing means Fifth calculation means for calculating the deviation, and a second control section for judging whether or not visual confirmation work for defects or foreign matters on the semiconductor wafer is necessary from the average value and standard deviation calculated by the fifth calculation means. And a warning means for warning a process abnormality based on the determination result of the second control means.

As a result, the appearance, defects output from the foreign substance inspection device, and abnormalities in the distribution of foreign substance data can be automatically monitored, and the process abnormality monitoring device can automatically determine whether or not a review is necessary. it can.

As described above, since it is automatically determined whether or not a review is necessary in the process of semiconductor wafer, the efficiency of the inspection process including the review can be improved.

[0030]

Embodiments of the present invention will be described below in detail with reference to the drawings.

FIG. 1 is a conceptual diagram showing a connection configuration of a process abnormality detection system according to an embodiment of the present invention, FIG.
3 is a block diagram of a process abnormality detection system according to an embodiment of the present invention, FIG. 3 is a flow chart diagram showing process steps of process analysis by the process abnormality detection system according to one embodiment of the present invention, and FIG. FIG. 5 is an explanatory diagram illustrating conversion from a chip coordinate system to a wafer coordinate system according to one embodiment of the invention, and FIG. 5 is an explanatory diagram showing division elements in a θ direction in a semiconductor wafer according to one embodiment of the present invention;
6 is an explanatory view showing the maximum radius of a dividing element in a semiconductor wafer according to one embodiment of the present invention, and FIG. 7 is an explanatory view showing a dividing element in the r direction in a semiconductor wafer according to one embodiment of the present invention, FIG. 8 is an explanatory view showing defect foreign matter density distributions in the r direction and the θ direction in a semiconductor wafer according to one embodiment of the present invention, and FIG. 9 is an explanatory view showing defective foreign matter distributions on a semiconductor wafer corresponding to FIG. FIG. 10 is an explanatory view showing defect foreign matter density distributions in the r direction and the θ direction in the case where there is no bias in the distribution of defective foreign matter density in a semiconductor wafer in one embodiment of the present invention, and FIG. FIG. 12 is an explanatory view showing the maximum and minimum values of the defect foreign matter density in the θ direction in the semiconductor wafer according to one embodiment, and FIG. 12 is a diagram showing the smoothness of the foreign matter defect density distribution in the θ direction in the semiconductor wafer according to one embodiment of the present invention. FIG. 13 is a diagram showing distributions before and after processing and after first-order differential processing, FIG. 13 is a flowchart of grouping processing according to one embodiment of the present invention, and FIG. 14 is a defect on a semiconductor wafer according to one embodiment of the present invention. FIG. 15 is a diagram showing an example of distribution of foreign matter, and FIG. 15 is a diagram showing another example of distribution of defects and foreign matter on a semiconductor wafer according to one embodiment of the present invention.

In the present embodiment, a process abnormality monitoring system (process abnormality monitoring device) 1 for monitoring a process abnormality of a semiconductor device includes an analysis system 2 for analyzing a process abnormality and a data storage means DB for storing coordinate data and the like. It is composed by.

Further, the process abnormality monitoring system 1 includes a visual inspection device GK for inspecting defects and shapes of a pattern formed on a semiconductor wafer, which will be described later, and a foreign substance inspection device PK for inspecting foreign substances on the semiconductor wafer. A review station RS for performing human visual confirmation of defects or foreign matter on a semiconductor wafer by looking at an enlarged image of a metal microscope or a scanning electron microscope for disconnection, short-circuit or foreign matter is connected via a network Nt. There is.

Further, the coordinate data output from the visual inspection device GK and the foreign substance inspection device PK are stored in the data storage means DB provided in the process abnormality monitoring system 1.

Next, the structure of the analysis system 2 will be described with reference to FIG.

First, the analysis system 2 is the appearance inspection apparatus G.
K, foreign matter inspection device PK and review station R
A communication control unit 3 that controls communication of data input / output from S is provided.

Then, the communication control section 3 uses the visual inspection apparatus G
K, foreign matter inspection device PK and review station RS
Is connected to the network Nt to which is connected.

The analysis system 2 also includes a communication control unit 3
A data input / output unit 4 serving as a data input / output unit for which communication control of data is performed is provided, and the data input / output unit 4 is connected to the communication control unit 3.

Further, the analysis system 2 is provided with a retrieval unit 5 for retrieving predetermined data from the coordinate data stored in the data storage means DB and a memory 6 for storing the coordinate data retrieved by the retrieval unit 5. Search unit 5
Are connected to the memory 6, and the memory 6 is connected to the data input / output unit 4.

The analysis system 2 is also provided with an input / output interface 7 for adjusting the timing of data exchange between the analysis system 2 and the data storage means DB. The search unit 5 and the data storage means DB are connected.

Further, the analysis system 2 performs a predetermined calculation based on a program storage unit 8 in which a program which is software for performing necessary processing in the process monitoring abnormality system 1 is stored and a program stored in the program storage unit 8. A computing unit (first to fifth computing means, dividing means) 9 for performing is provided.

Further, the analysis system 2 is provided with an alarm generator (warning means) 10 for notifying an abnormality by a warning sound or a warning light when a process is abnormal, and a display section 11 for displaying data.

Further, the analysis system 2 is provided with a main control section (first and second control means) 12 which controls all the control in the analysis system 2, and the main control section 12 has data input / output. Section 4, search section 5, program storage section 8
Also, the alarm generator 10 is connected.

The arithmetic unit 9 is connected to the memory 6, and the display unit 11 is connected to the data input / output unit 4.

Next, the operation of this embodiment will be described.

First, the appearance inspection apparatus GK and the foreign matter inspection apparatus P
Coordinate data of defects and particles detected by K
Transfer to the analysis system 2 via t.

At this time, the coordinate data of defects and foreign particles is given as XY coordinate data for each semiconductor chip. The analysis system 2 performs the processing shown in the flowchart of FIG. 3 on this data.

First, the retrieval unit 5 retrieves predetermined coordinate data stored in the data storage means DB, reads it into the memory 6 (step S500), and the arithmetic unit 9 based on the coordinate data stored in the memory 6. Performs calculation to convert the XY coordinate data for each semiconductor chip into wafer coordinate data having the center of the semiconductor wafer as the origin (step S50).
1).

Now, the process of step S501 will be described with reference to FIG. For example, in the semiconductor wafer W,
The coordinates (X, Y) of the defect 61, the foreign matter, etc. in the semiconductor chip CH1 in the m-th row and the n-th column are the wafer coordinates (Xw, Y) by the following equation.
w).

[0050]

[Equation 1]

[0051]

[Equation 2]

However, Px and Py are chip pitches in the X and Y directions, and m0 and n0 are semiconductor chips CH which are origins.
The row and column numbers 2, Δx and Δy, indicate the distances between the chip origin 601 of the semiconductor chip CH2 and the wafer center 600 in the X and Y directions.

Next, with respect to the wafer coordinates (Xw, Yw), the position data in the r (radius) direction and the position data in the θ (circumferential) direction, which are one-dimensional data, are obtained by the following equations, and the polar coordinates (r, θ) are obtained. Is converted to (step S502).

[0054]

(Equation 3)

[0055]

(Equation 4)

Further, the defect foreign matter density P, which is the density of defects and foreign matter of each divided element i, j in the r direction and the θ direction
ri and Pθj are obtained (step S503).

As shown in FIG. 5, the dividing element i in the r direction is a circle when i = 1 and a ring zone when i> 1.
The area Sri of each element i is

[0058]

(Equation 5)

[0059]

(Equation 6)

It can be represented by However, Δr is

[0061]

(Equation 7)

And rmax is the distance between the wafer center 600 and the orientation flat 602 (hereinafter referred to as the orientation flat), as shown in FIG. 6, for example. Further, imax is the number of divisions in the r direction and is set to an arbitrary value.

Here, assuming that the number of defective foreign matters, which is the number of defects and foreign matters contained in the dividing element i, is Nri, the defective foreign matter density Pri is

[0064]

(Equation 8)

It becomes

On the other hand, the divided element j in the θ direction has a fan shape having the same area as shown in FIG. 7, and the area of each element j is

[0067]

[Equation 9]

Is obtained. However, jmax is the number of divisions in the θ direction, and an arbitrary value is set.

Here, the number of defective foreign matters in the divided element j is Nθ.
j, the defect particle density Pθj is

[0070]

(Equation 10)

Is obtained.

Therefore, as a result of the processing in step S503, the defects in the r direction as shown in FIG. 8A and the defect foreign substance density distribution which is the density distribution of the foreign substances and the defects in the θ direction as shown in FIG. 8B are obtained. A foreign matter density distribution can be obtained.

These are semiconductor wafers W shown in FIG.
Corresponds to the upper defect particle distribution KB.

Next, the abnormality of these distributions is checked by the main control unit 1.
2 determines (step S504). For example, assuming that the defect foreign matter density has no distribution bias, the defect foreign matter density distributions in the r and θ directions are constant values Pra and Pθa as shown in FIGS. 10 (a) and 10 (b). Where Pr
The calculation formulas for a and Pθa are shown.

[0075]

[Equation 11]

[0076]

(Equation 12)

Then, in order to test the anomaly of the distribution using, for example, the percent points of the chi-square distribution, the statistics Cr and Cθ in the r and θ directions are calculated by the following equations.

[0078]

(Equation 13)

[0079]

[Equation 14]

For example, the assumption that "the distribution of defective foreign matter density has no abnormality" is denied, that is, the distribution is determined to be abnormal, the following equation may be established in the r and θ directions.

[0081]

(Equation 15)

[0082]

(Equation 16)

Here, CS represents the percentage point of the chi-square distribution and is a function of the rejection rate α and the degree of freedom ν.

The CS value is a general textbook of statistics, for example, November 21, 1994, published by The University of Tokyo Press, Nozomu Matsubara et al. (Introduction to Statistics) P282-P
283.

For example, if the probability that the above judgment is erroneous is 5% in each of the r and θ directions, α1 = α2 =
It will be 0.05. The degrees of freedom ν1 and ν2 are

[0086]

[Equation 17]

[0087]

(Equation 18)

It is shown by.

Next, as another means for judging the abnormality of the distribution, as shown in FIG. 11, the defect particle density P in the θ direction is shown.
Taking the distribution of θj as an example, the maximum value Pθmax and the minimum value P
A contrast value CNT that can be expressed by the following equation is obtained from θmin.

[0090]

[Equation 19]

Here, if the contrast value CNT is larger than the preset threshold value Cth, the main control unit 12 determines that it is abnormal.

If it is determined in step S504 that the distribution is not abnormal, the process proceeds to step S510, and if it is determined to be abnormal, the process proceeds to step S505.

For example, if it is determined that the distribution is not abnormal, it is determined whether the average defect foreign matter densities Pra, Pθa exceed a preset threshold value (step S5).
10) If it exceeds the threshold, the main control unit 12 outputs a predetermined signal to the alarm generator 10 to generate an alarm or turn on a warning light to warn the operator that it is a review target (step S508). ).

If the threshold value is not exceeded, the process returns to step S500 to read the next coordinate data.

Next, if it is determined in step S504 that there is an abnormality in the distribution, the distribution is clustered, that is, grouped (step S505).

Here, the distribution of the defect foreign matter density Pθj in the θ direction shown in FIG. 12A will be taken again as an example. The distribution of the defect foreign matter density Pθj is smoothed as a pre-process for performing grouping that does not depend on the local defect foreign matter density.

The smoothing method may be a moving average method in which each defect foreign matter density is replaced by an average value including several neighboring points, or after Fourier transform, a low pass filter is applied and inverse Fourier transform is performed. A well-known method may be applied to the signal processing of. FIG. 12B shows the distribution after smoothing.

Next, in order to determine the boundary for grouping, the first derivative of Pθj is obtained. Now, FIG.
Since the horizontal axis j of the smoothed distribution shown in (b) is an integer, P
The first derivative dPθj of θj with respect to j is, for example,

[0099]

(Equation 20)

[0100]

(Equation 21)

The distribution of dPθj obtained by the above equation is shown in FIG. 12 (c).

Here, in FIG. 12C, dPθj
Points B1 and B2 at which is changed from negative to positive are defined as boundaries between groups (clusters).

As a result, the distribution shown in FIG. 12B is divided into group A1 and group A2. Assuming that FIG.
If the distribution in (b) is the distribution in the r direction, A1 divided by B2 becomes another group A3, which is divided into three groups in total.

Here, the flow of the grouping process is shown in FIG.
It is shown in the flowchart of FIG.

First, the distribution is smoothed (step S55).
1), after calculating the first derivative of the distribution (step S552), dPθj · dPθj + 1 <0 and dPθj + 1> 0
Then, j is obtained (step S553), j is taken as the boundary point of the group (step S554), and j and j + 1 are taken as the boundary of the group (step S555).

After the grouping, the group number of each group is set to k (k = 1 to K, where K is the number of groups), and in step S506 of FIG. 3, the average values μrk, μθk and standard deviation of each group are set. Calculate σrk and σθk.

For example, if the region of the group number 1 in the r direction is i = i1 to imax1, the average value μr1 and the standard deviation σr1 are

[0108]

(Equation 22)

[0109]

(Equation 23)

It is calculated by

The average value μθk in the θ direction and the standard deviation σθk can be calculated in the same manner as the above equation. These parameters K, μr
The distribution of defects and foreign substances can be quantitatively expressed by k, μθk, σrk, and σθk.

For example, in the case of the distribution KB as shown in FIG. 14, K = 2, μr1 = 0.5, σr1 = 0.1, μθ1 =
0.25, σθ1 = 0.25, μr2 = 0.5, σr2 = 0.
1, μθ2 = 0.75 and σθ2 = 0.1. Where r
The 1 in the direction indicates rmax, and the 1 in the θ direction indicates 360 degrees.

In the case of the distribution KB shown in FIG. 15, K
= 1, μr1 = 0.5, σr1 = 0.4, μθ1 = 0.25,
σθ1 = 0.05.

In step S507 of FIG. 3, the shapes of the distributions of defects and foreign matters to be reviewed are previously set to K, μrk, μ.
If the main control unit 12 determines that the distribution is represented by the inequalities of θk, σrk, and σθk, and the distribution is applicable to this, the main control unit 12 outputs a predetermined signal to turn on a warning light, an alarm, a screen display, or the like. At step S508, the main control unit 12 outputs the coordinate data from the communication control unit 3 via the data input / output unit 4 and transfers the coordinate data to the review station RS via the network Nt. A review is performed (step S509).

Further, K, μr obtained in step S506
k, μθk, σrk, and σθk are registered in the data storage means DB for each semiconductor wafer W.

On the other hand, if the conditional expressions of the above parameters are used, it is possible to retrieve data having a similar distribution from the past data stored in the data storage means DB of the process abnormality monitoring system 1 in FIG. You can

As a result, it is possible to efficiently analyze the current data and take countermeasures by referring to the failure analysis performed on the past data and the countermeasures.

Further, a system for inputting the above parameters and outputting the cause of failure by fuzzy reasoning or a neural network by making the relation between these parameters and the cause of failure into rules or by having a computer learn. It may be added to the analysis system 2 of FIG.

By the way, the semiconductor device is manufactured by repeating film formation, exposure and etching in each step, and the inspection is also performed in each step.

Here, FIG. 16 shows a transition of the number of defective foreign matters detected in each step examined by the present inventor in each step. The defect and foreign matter in the previous steps A and B are also in step C. It can be seen that it will be detected again. Therefore, as the process progresses, the number of defects and foreign substances already reviewed in the defects and foreign substances to be reviewed increases, and it is understood that time efficiency is poor.

Therefore, as shown in FIG. 17, the defect foreign matter coordinate data, which is the coordinate data of the defect and foreign matter in the current process on the semiconductor wafer W, is stored in the data storage means DB of the process abnormality monitoring system 1 of FIG. Substituting the defect particle coordinate data of the process before
A system for obtaining defective foreign matter coordinate data newly found in the current process may be added to the analysis system 2 in FIG.

In step S509 in FIG. 3, if defective foreign matter coordinate data newly found in the current process is transferred to the review station RS, the duplication with the previous review is eliminated and the efficiency of the review work can be improved.

Although the invention made by the present inventor has been specifically described based on the embodiments of the present invention, the present invention is not limited to the above embodiments and various modifications can be made without departing from the scope of the invention. It goes without saying that it can be changed.

For example, in the above-described embodiment, after converting the wafer coordinates into polar coordinates, the defect foreign matter density, which is the density of the defects and foreign matter of each divided element, is obtained. Direction and length (Y)
Even if the defect foreign matter density distribution is calculated by calculating the defect foreign matter density which is the density of the defect and foreign matter of each dividing element after converting into the position data which is the one-dimensional data of the direction, the process abnormality can be monitored well. You can

[0125]

Advantageous effects obtained by typical ones of the inventions disclosed by the present application will be briefly described as follows.
It is as follows.

(1) According to the present invention, the process abnormality monitoring device automatically monitors the defect in the semiconductor wafer and the abnormality in the distribution of the foreign substance data, and automatically determines whether or not the review is necessary. Process anomaly detection can be automated.

(2) Further, according to the present invention, since the distribution of defect and foreign particle data on the semiconductor wafer can be searched, the defect analysis based on the similar distribution data can be performed more efficiently.

(3) Further, in the present invention, the inspection efficiency in the inspection process of the semiconductor wafer including the reviews can be improved by the above (1) and (2),
The manufacturing efficiency of semiconductor devices can also be increased.

[Brief description of the drawings]

FIG. 1 is a conceptual diagram showing a connection configuration of a process abnormality detection system according to an embodiment of the present invention.

FIG. 2 is a block diagram of a process abnormality detection system according to an embodiment of the present invention.

FIG. 3 is a flowchart showing process steps of process analysis by the process abnormality detection system according to the embodiment of the present invention.

FIG. 4 is an explanatory diagram illustrating conversion from a chip coordinate system to a wafer coordinate system according to an embodiment of the present invention.

FIG. 5 is an explanatory diagram showing divisional elements in the θ direction in the semiconductor wafer according to the embodiment of the present invention.

FIG. 6 is an explanatory diagram showing a maximum radius of a dividing element in a semiconductor wafer according to an embodiment of the present invention.

FIG. 7 is an explanatory diagram showing division elements in the r direction in the semiconductor wafer according to the embodiment of the present invention.

FIG. 8 is an explanatory diagram showing defect foreign matter density distributions in the r direction and the θ direction in a semiconductor wafer according to an embodiment of the present invention.

9 is an explanatory view showing a defect foreign matter distribution on a semiconductor wafer corresponding to FIG.

FIG. 10 shows r in the case where there is no uneven distribution of defect foreign matter density in a semiconductor wafer in one embodiment of the present invention.
It is explanatory drawing which shows the defect foreign material density distribution of a direction and (theta) direction.

FIG. 11 is an explanatory diagram showing the maximum value and the minimum value of the defect foreign matter density in the θ direction in the semiconductor wafer according to the embodiment of the present invention.

FIGS. 12A and 12B are a diagram illustrating a case of smoothing a foreign matter defect density distribution in a θ direction on a semiconductor wafer according to an embodiment of the present invention before and after a smoothing process
It is a figure which shows distribution after a secondary differential process.

FIG. 13 is a flowchart of a grouping process according to the embodiment of the present invention.

FIG. 14 is a diagram showing an example of distribution of defects and foreign matter on a semiconductor wafer according to an embodiment of the present invention.

FIG. 15 is a diagram showing another example of distribution of defects and foreign matters on a semiconductor wafer according to one embodiment of the present invention.

FIG. 16 is an explanatory diagram showing the number of defects in each process examined by the inventor and a breakdown of defect generation processes.

FIG. 17 is a schematic diagram showing a map on a semiconductor wafer of a current process defect, a previous process defect, and a review target defect after subtraction processing thereof in another embodiment of the present invention.

[Explanation of symbols]

1 Process Abnormality Monitoring System (Process Abnormality Monitoring Device) 2 Analysis System 3 Communication Control Section 4 Data Input / Output Section 5 Search Section 6 Memory 7 Input / Output Interface 8 Program Storage Section 9 Arithmetic Section (1st to 5th Arithmetic Means, Division) 10) Alarm generator (warning means) 11 Display unit 12 Main control unit (first and second control unit) 61 Defect DB data storage unit GK Appearance inspection device PK Foreign matter inspection device RS review station Nt network W Semiconductor wafer CH1 Semiconductor chip CH2 Semiconductor chip KB distribution 600 Wafer center 601 Chip origin 602 Orientation flat AB process

Claims (9)

[Claims] 1. A step of measuring a distribution state of defects and foreign matters on a semiconductor wafer, a step of converting the measured distribution state into at least one one-dimensional data,
Dividing the converted one-dimensional data into arbitrary dividing elements, calculating a defect foreign matter density which is a density of defects and foreign matter in the dividing elements, and calculating the calculated defect foreign matter density in the dividing elements. On the basis of the defect and foreign matter density distribution, there is no abnormality in the defective foreign matter density distribution, a step of grouping the one-dimensional data, and a defective foreign matter density of each grouped group. And the step of calculating the average value and standard deviation of the number of defects and the number of foreign particles for each grouped group, and the visual confirmation of defects or foreign particles of the semiconductor wafer from the calculated average value and standard deviation And a step of determining whether or not work is required. 2. The process abnormality monitoring method according to claim 1, wherein the one-dimensional data is position data in a circumferential direction, a radial direction, or a horizontal direction or a vertical direction of the semiconductor wafer. Monitoring method. 3. The process abnormality monitoring method according to claim 1, wherein the grouping step includes a step of smoothing the defect foreign matter density distribution, and a step of calculating a first derivative of the defective foreign matter density. And a step of obtaining a characteristic point of the distribution of the first derivative of the defect foreign matter density. 4. The process abnormality monitoring method according to claim 3, wherein the characteristic point of the distribution of the first derivative of the defect particle density is that the value of the first derivative of the defect particle density changes from negative to positive. A process abnormality monitoring method characterized by the above. 5. The process abnormality monitoring method according to claim 1, wherein the step of determining whether or not there is an abnormality in the defect foreign matter density distribution includes the defect foreign matter density and the defect foreign matter density. A process abnormality monitoring method comprising a step of comparing a value obtained by dividing a square of a difference from an average value by an average value of defect foreign matter density and a percentage point of a chi-square distribution. 6. The process abnormality monitoring method according to claim 1, wherein the step of determining whether or not there is an abnormality in the defect foreign matter density distribution includes a maximum value and a minimum defect foreign matter density. A process abnormality monitoring method comprising a step of comparing a contrast value obtained from the value with a reference value of a preset contrast value. 7. The process abnormality monitoring method according to claim 1, wherein the step of determining whether or not visual confirmation work for defects or foreign matter on the semiconductor wafer is required is an average value and a standard value. A process abnormality monitoring method comprising a step of comparing a deviation with a preset threshold value. 8. The process abnormality monitoring method according to claim 7, wherein the step of determining whether or not the visual confirmation work for defects or foreign matter on the semiconductor wafer is required is an average of the number of defects and the number of foreign matters on the entire semiconductor wafer. A process abnormality monitoring method comprising a step of comparing a value with a preset threshold value. 9. A data storage unit for storing measurement data of a distribution state of defects and particles on a semiconductor wafer, and at least one distribution state stored in the data storage unit.
First computing means for converting into one one-dimensional data, dividing means for dividing the one-dimensional data converted by the first computing means into arbitrary dividing elements, and the dividing element divided into the dividing means Second calculation means for calculating a defect foreign matter density, which is the density of defects and foreign matter inside, and a defect foreign matter density which is a distribution of defects and foreign matter density based on the defect foreign matter density calculated by the second calculation means The first control means for determining whether or not there is an abnormality in the distribution, the third calculation means for grouping the one-dimensional data converted by the first calculation means, and the third calculation means Fourth computing means for calculating the defective foreign matter density for each grouped group, and calculating the average value and standard deviation of the number of defects and the number of foreign matter for each grouped by the third computing means A fifth calculation unit, a second control unit that determines whether or not visual confirmation work for defects or foreign substances on the semiconductor wafer is necessary from the average value and standard deviation calculated by the fifth calculation unit, and A process abnormality monitoring device comprising: a warning unit that issues a process abnormality warning based on the determination result of the second control unit.