JP2002202806A - Process control system and manufacturing method for article - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a process control system which can perform proper process control and the quality control of an article even when set values of processing parameters of a process frequently vary, and a manufacturing method for the article. SOLUTION: Characteristic values which are related by kinds are stored and means values of characteristic values by the kinds which are related by the kinds and standard deviation are computed. A specific kind scheduled to be processed is selected and inputted and set values of processing parameters are determined; and the processing is performed according to the processing parameters and the obtained characteristic values are standardized by using the means values and standard deviation. The mean values and range of the standardized characteristic values of plural-time processing are computed and it is judged whether or not a process is abnormal according to the mean value and range of the standardized characteristic values.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a process management system and a method for manufacturing an article, and particularly to a case where the set values of process parameters of a process frequently change as in the case of high-mix low-volume production. The present invention also relates to a process management system and an article manufacturing method capable of performing appropriate process management and article quality control.

[0002]

2. Description of the Related Art In recent years, when mass-produced products typified by semiconductor devices are manufactured, an X control chart, Xba
Create a control chart such as an r-R control chart, determine whether the characteristic values of the products and semi-finished products included in a certain lot are the desired values, and monitor whether the manufacturing process is in a statistical control state A technique is used. Generally, a normal distribution, a binomial distribution, or the like is used as a model for creating such a control chart.

[0003]

The statistical management state is
"It is a state where almost all the points hit on the control chart are within the control limit value." If it is in a stable state, the only cause of the change is accidental change, and there is no cause that cannot be overlooked. (Kaoru Ishikawa: "Control Chart", see Nikka Giren Publishing Company (1962)).

However, such a control chart monitors a change in a characteristic value under the same set value of a processing parameter in a manufacturing process. For this reason, in an actual product manufacturing process, if the set values of the processing parameters in each manufacturing process are frequently changed by various types, the products cannot be continuously used.

[0005] In such a state, it is not easy to find an abnormality in the manufacturing process and its cause, so that a defective product is produced due to a delay in finding the defect, and the efficiency of the defect analysis is reduced. Further, if such a state continues, there is a possibility that the management of the process using the control chart at the manufacturing site may be stopped.

It is an object of the present invention to provide a process management system and an article manufacturing method capable of performing appropriate process management and article quality control even in the case of high-mix low-volume production.

[0007]

SUMMARY OF THE INVENTION A process management system according to the present invention is a system for managing a process having a process common to a plurality of product types. Storage means for storing a characteristic value associated with each type by inputting a plurality of subsequent characteristic values for each type;
A first calculating means for calculating an average value and a standard deviation associated with each kind of the characteristic value of each kind based on the data stored in the storage means, and one specific kind to be processed in the process Type input means for selecting and inputting from among the plurality of types, setting value determining means for determining a set value of a processing parameter corresponding to a type determination signal from the type input means, and setting value of the processing parameter The characteristic values after processing of the specific type corresponding to the type determination signal processed based on the above are standardized using the average value and the standard deviation of the characteristic values associated with the specific type calculated by the first calculating means. A second calculating means for calculating the average value and range of the standardized characteristic values obtained by the second calculating means for a plurality of processes, and a standardized characteristic value Based on the average value and the range, in which comprises an abnormality determination means for determining whether the process abnormality, and the abnormality determination means determines that there is abnormality and warning means for issuing a warning, the.

A method of manufacturing an article according to the present invention is a method of manufacturing an article having a common process for a plurality of types, and for a certain processing parameter in the common process, the characteristic value after the processing of each type is individually set. Storing the characteristic values associated with each of the varieties by inputting a plurality of the varieties, and associating the characteristic values of each of the varieties with the respective varieties based on the characteristic values associated with the respective varieties. Calculating the calculated average value and standard deviation, selecting and inputting one specific product to be processed in the above-described process from the plurality of products, and setting processing parameters corresponding to the specific product. Determining a value, and associating the processed characteristic value of the specific type processed based on the set value of the processing parameter with the specific type. Standardizing using the average value and the standard deviation of the obtained characteristic values, calculating the average value and range of the standardized characteristic values obtained in the step of standardizing a plurality of processes, and Judging the presence or absence of an abnormality in the above process based on an average value and a range of the values.

Further, the process is a film forming process of an electronic component,
The processing parameter is a film formation time and / or a film formation rate, and the characteristic value is a film thickness.

The process is a photolithography process of an electronic component, the processing parameter is an exposure time, and the characteristic value is a line width of a wiring pattern.

The process is an electronic component etching process, the processing parameter is an etching time and / or an etching rate, and the characteristic value is a line width of a wiring pattern.

Further, the line width of the wiring pattern can be obtained by measuring the line width of the evaluation wiring pattern formed of the same pattern regardless of the type.

[0013]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiment 1 A method for manufacturing an article according to Embodiment 1 of the present invention will be described. In certain process P that performs processing common for a plurality of varieties, m types of processing parameters X _1, X _2, · · ·, set values of X _m are different depending on the variety, the n type by treatment with step P Consider a case where characteristic values Y ₁ , Y ₂ ,..., Y _n are obtained. FIG. 1 is a diagram illustrating a method for manufacturing an article according to Embodiment 1 of the present invention. Processing parameters X ₁ , X ₂ ,... From past processing data of the process P or processing data of the prototype.
- the average value mu ₁ of X _m and the characteristic value, mu _2, · · ·, the relationship between the mu _n as functions f _j, also process parameters X _1, X _2, ·
· ·, X _m and the standard deviation σ _1, σ _2, ···, obtained in advance the relationship between the sigma _n as a function g _j (step S1). The function f _j and the function g _j can be obtained by regression analysis from a database of the relationship between the set values and the characteristic values of the processing parameters obtained in the past processing or the prototype processing. Accordingly, it is possible to cope with a case where the set value of the processing parameter changes depending on various types. μ _j = f _j (X ₁ , X ₂ ,..., X _m ) (1 ≦ j ≦ n) (1) σ _j = g _j (X ₁ , X ₂ ,..., X _m ) ( 1 ≦ j ≦ n) (2)

A specific type to be processed in the process P at a certain time is determined, and m types of processing parameters X ₁ , X ₂ ,
., Enter a set value corresponding to that particular varieties X _m (step S2). In this process P, processing is performed based on the processing parameters, and n types of characteristic values Y ₁ , Y ₂ ,.
· To obtain Y _n (step S3). Next, the characteristic values Y ₁ , Y
_2, ..., standardized characteristic Y _n by the following equation (3) value K _j
(Step S4). The average value μ ₁ , μ ₂ ,.
· ·, Mu _n and the standard deviation σ _1, σ _2, ···, the sigma _n used as obtained in step S1. This conversion is called standardization of characteristic values, and thereby, it is possible to offset a difference in characteristic values due to a difference in set value depending on a product type.

(Equation 1)

As the Z-th data, n kinds of K _j (1
≦ j ≦ n) (step S5). Step S2
By repeating steps S5 k times from obtaining k number of K _j successive to chronological order.

Next, the following equation (4) and the average value Xbar _j and scope R _j (1 ≦ _j ≦ standardized characteristic value K _j by Equation (5)
n) (step S6). That mean Xbar _j sums the k pieces of K _j consecutive in chronological order, divided by k, √k
Multiply by Moreover, the range R _j multiplies the √k to what maximum value among the k K _j from (K _jmax) minus minimum value of the k K _j a (K _jmin). The reason for multiplying by √k is
This is a correction term for managing the distribution of the aggregate values of k pieces of observation data, and is known as a central limit theorem.

(Equation 2)

(Equation 3)

The average value Xbar _j and the range R _j obtained in step S6 are plotted on the control chart C _j (1 ≦ j ≦ n) (step S7). The control chart _Cj will be described. FIG. 2 is a diagram showing an example of an Xbar _j control chart of the present invention, and FIG. 3 is a view showing an example of an R _j control chart of the present invention. The horizontal axis of the control chart _Cj is a time series, and is plotted in the order in which the processing is performed in the process P. The control chart C _j is a control chart for managing the standardized characteristic value K _{j and} the characteristic value Y _j before being standardized.
As can be seen from the above equation (3), when the characteristic value Y _j is equal to the average value μ _j , K _j = 0, and when Y _j is ± 3σ _j apart from μ _j , K _j = ± 3. Becomes Average Xbar _j and R _j K _j, to those taking range, since it is multiplied by the √k, upper control limit value Xbar _j control chart (UC
L) = + 3 and lower control limit (LCL) = − 3,
By providing the R _j management upper control limit value in view (UCL) = + 3, it is possible regardless of variations in the change of the set value of the characteristic value, observed in the same conditions. Here, Xbar _j and R _j are the average of K _j , the range obtained by multiplying the average and the range by K, and the upper control limit (UCL) of the control chart C _j = + 3 / √k, and the lower control limit. (LCL) = − 3 / √k.
However, in this case, when the value of k changes in the middle for some reason, the vertical axis of the control chart must be recreated, and when judging the process state from the control chart when the control limit value is ± 3. In addition, it is difficult to intuitively judge whether or not the value exceeds the control limit value only by looking at the value of Xbar. Accordingly, the average value of K _j as Xbar _j and R _j as described above, to those taking range, it is more desirable to multiplying the previously √k. The control limit value is not limited to ± 3, but may be ± 2, for example, and can be set as appropriate.

Next, the control chart C _j on which Xbar _j and R _j are pointed is monitored to determine whether or not the process P is in a statistical management state (step S8). That is, if there is no change other than the random error in step P, it is dotting the control chart C _j Xb
No abnormality is found in which ar _j and R _j exceed the control limit value or cause a peculiar habit. If no abnormality is found, the process returns to step S2, and the processes after step S2 are repeated. For some missed not cause contrary Further, when a change in process P has occurred, Xbar _j and R _j, which are dotting the control chart C _j is, or exceeds the control limit value, the abnormality, such as specific habit or cause Is found, the process P is determined to be abnormal, and the abnormality is notified (step S9). With the method as described above, the abnormality in the process P can be detected promptly, and the cause can be investigated and countermeasures can be taken. As a result, it goes without saying that the production of defective products is reduced and the cost of countermeasures is reduced.
The method for manufacturing an article of the present invention as described above can be used for manufacturing all kinds of articles in addition to manufacturing electronic parts. Further, the process P may be a manufacturing process for performing a processing process, or an inspection process which is a non-processing process,
Any process can be used as long as a characteristic value can be obtained corresponding to the processing parameter.

Embodiment 2 Embodiment 2 of the present invention will be described. The second embodiment is an apparatus obtained by systematizing the article manufacturing method of the first embodiment. FIG.
FIG. 6 is a diagram illustrating a configuration of a process management system according to a second embodiment of the present invention. Hereinafter, the configuration of the process management system will be described with reference to FIG. The process management system according to the second embodiment of the present invention is a system that manages a process P8 including a process common to a plurality of types, a characteristic value storage unit 1 that stores characteristic values associated with each type,
A kind input means 2 for determining a plurality of kinds for one specific kind to be processed in the process P8 and inputting the kind data 3, and setting values of processing parameters corresponding to a kind determination signal from the kind input means 2 And a mean value and standard deviation (μ _j , σ _j ) 6 based on the selected set value 4 and the data stored in the characteristic value storage means 1.
And a mean value / standard deviation calculating means 7 for obtaining the characteristic value (Y _j ) 9 and the obtained characteristic value (Y _j ) 9 are converted into the average value μ _j and the standard deviation A standardized characteristic value calculating means 11 for obtaining a standardized characteristic value (K _j ) 10 calculated from σ _j using the above equation (3), and the standardized characteristic value (K _j ) 10 are stored. The standardized characteristic value storage means 12 for counting and the standardized characteristic value (K _j ) 10 which is processed a plurality of times in the process P8 and stored when the stored standardized characteristic value reaches a predetermined number, are calculated by the above equations (4) and (5). ) Using the standardized characteristic value (K _j ) 1
Average value / range calculating means 14 for calculating the average value and range (Xbar _j , R _j ) 13 of 0, and the calculated average value and range 13
, A control chart display means 15 for plotting the values on the control chart and displaying it as a control chart, an abnormality determining means 16 for determining whether or not there is an abnormality from the average value and the value of the range 13,
And warning means 17 for issuing a warning when it is determined in step 6 that there is an abnormality.

The process management system according to the second embodiment will be described in more detail. In the characteristic value storage means 1, various setting values are input to the processing parameter a plurality of times, and characteristic values obtained by performing processing a plurality of times based on the various setting values are stored in association with the setting values. . As the various set values, set values corresponding to all the types processed in the process P8 may be used, but it is not always necessary to use the set values corresponding to all the types. Is determined by regression analysis as a function represented by the above equations (1) and (2).
A characteristic value associated with each product type may be stored as such a function. The characteristic value storage means 1 may be a storage device of a computer, or may be written by a worker in a table or the like. The type input means 2 is an input device for selecting a specific type to be processed in the process P8 and inputting the type data 3 thereof. For example, it is a keyboard of a computer terminal, a barcode reader, a selection switch, or the like.

The set value determining means 5 selects a unique set value 4 corresponding to the specific kind set by the kind input means 2. The selection of the setting value 4 is performed by referring to a table (database) listing pairs of product types and setting values created in advance. The operation up to this point may be manually performed by the operator at the time when the specific type is determined, by referring to a type-to-set value correspondence table or the like.

In the process P, m kinds of processing parameters X ₁ , X ₂ ,..., X _m are prepared. Although the number m of processing parameters depends on the process, it is preferable that the number m be small in consideration of the management and operation load. The set value 4 of the processing parameter is determined by the average value / standard deviation calculating means 7 and the process P
8 is input.

When the set values 4 of m kinds of processing parameters are inputted to the average value / standard deviation calculating means 7, n kinds of processing parameters are stored based on the characteristic values associated with each product type stored in the characteristic value storing means 1. And the standard deviation (μ _j , σ _j ) 6 are obtained. Usually, the average value and the standard deviation (μ _j , σ _j ) 6 are obtained by the functions shown in the above equations (1) and (2). The average value / standard deviation calculation means 7 may be programmed software, or may be manually calculated by an operator.

On the other hand, when a set value 4 of m kinds of processing parameters is input to the process P8, the set value 4 is set in the process P8.
, And n types of characteristic values (Y _j ) 9 are obtained.

Average value μ _j , standard deviation σ _j and characteristic value Y
_j is input to the standardized characteristic value calculating means 11 and is converted into n kinds of standardized characteristic values (K _j ) 10 by the above equation (3). The standardized characteristic value calculation means 11 may be programmed software, or may be manually calculated by an operator.

The converted standardized characteristic value (K _j ) 10 is stored in the standardized characteristic value storage means 12. The standardized characteristic value storage means 12 may be a storage device of a computer, or may be written in a table by an operator.

The standardized characteristic value storage means 12 includes a counter for counting the number of sets of the stored standardized characteristic values _Kj . A step of inputting the type data to the type input means 2 when the number k of sets of the standardized characteristic values K _j to be accumulated is set in advance and the number of sets of the standardized characteristic values K _j is less than the set number k. And the same processing is repeated. When the number of sets of the standardized characteristic values K _j reaches k, the accumulated k standardized characteristic values K _j are input to the average value / range calculating means 14.

K input to the average / range calculating means 14
The standardized characteristic value K _{j of the set} is calculated by using the above-described equations (4) and (5) to obtain n kinds of average values and ranges (Xbar _j , R _j ) 13.
Is converted to The average / range calculation means 14 may be programmed software, or may be manually calculated by an operator.

The control chart display means 15 places and displays the values on n kinds of Xbar control charts and R control charts based on the n kinds of average values and the values of the range (Xbar _j , R _j ) 13. . FIG. 2 shows an example of the Xbar control chart, and FIG. 3 shows an example of the R control chart.
Shown in The control chart display means 15 may be a display monitor of a computer, or may be an operator who has plotted on graph paper.

On the other hand, the abnormality judging means 16 judges whether or not the process P8 is in a stable state based on the n kinds of average values and the values of the range (Xbar _j , R _j ) 13. Normally the value of Xbar _j is ±
When 3 is exceeded, or, an abnormality is determined when the value of R _j exceeds +3, or displays a message to this effect on the warning means 17 generates an alarm sound. On the other hand, when the abnormality determination unit 16 determines that the state is stable, the operation is repeated from the operation of setting the type to be processed in the type input unit 2. Abnormality determination means 16
May be programmed software or may be performed by an operator. Further, the warning means 17 may be a display monitor of a computer or a buzzer for generating an alarm sound. Further, ordinary communication means (oral communication, telephone, document transmission, e-mail, etc.) by the operator may be used.

In the process management system according to the present invention, it is possible to know in real time whether or not the process P is in a statistical management state even in the case of high-mix low-volume production without delay. The effect is described in the first embodiment.
Needless to say, this is the same.

Embodiment 3 FIG. Embodiment 3 of the present invention will be described. In the third embodiment, the process is a film forming process of an electronic component. In this embodiment, a case where a wiring metal film is formed on a wafer of a surface acoustic wave device (hereinafter, referred to as a SAW device) will be described as an example. Also, there is no difference in effect.

In the wafer process of the SAW device, a process of forming a metal thin film mainly composed of aluminum (Al) on the surface of a piezoelectric material wafer such as lithium tantalum oxide (LiTaO ₃ ) or lithium niobium oxide (LiNbO ₃ ) is performed. There is. As a method of forming a metal thin film, a method such as a sputtering method, a chemical vapor deposition method (CVD), or a vapor deposition method can be used, and any method may be used in the operation of the present invention.

The SAW device is a kind of band-pass filter, and its frequency characteristic differs depending on the type. Also, it is widely known that one of the important factors that determines the frequency characteristics of a SAW device is the thickness of the wiring, that is, the thickness of the metal thin film on the wafer (for example, edited by IEICE, K. Shibayama). Supervision “Surface acoustic wave engineering” p.84). This indicates that the thickness of the metal thin film to be formed differs depending on the product type, and at the same time, in order to obtain a metal thin film having a target film thickness, it is necessary to perform process control of the film forming process. Indicates what must be done.

When the same kind is continuously processed in large quantities, it is sufficient to control the thickness of the metal thin film formed under the same film forming conditions by the conventional method. Since the target film thickness itself varies depending on the product type, there has been a problem in managing the film forming process.

In this embodiment, a sputtering method is used as a film forming method, and a film forming time X is used as a processing parameter.
₁ [Second] was selected (m = 1). Since it is only necessary to select a factor to be adjusted as a factor for controlling the film thickness for each product type, the film forming time is fixed, and for example, a film forming rate [nm / sec], a sputtering accelerating electric field [V], and the like are appropriately set as processing parameters. You may choose. Further, a plurality of factors may be selected, such as selecting both the film formation time and the film formation rate as processing parameters.
In these cases, a similar effect can be obtained.

In this embodiment, the thickness (n = 1) of the metal thin film is selected as the characteristic value Y ₁ . In implementation, the film thickness at the center of the wafer was measured for each wafer. Since the frequency of measurement also depends on the stability of the target film forming process, the film thickness at a plurality of coordinates (for example, 13 points in the plane) on the wafer surface may be used as the characteristic value for more strict control. For simpler management, the film thickness of one wafer may be measured for each film formation batch. Note that a fluorescent X-ray type film thickness measuring device was used for measuring the film thickness of the metal thin film.

The mean value / standard deviation calculating means 7 will be described. The relationship between the film forming time X _{1 as} a processing parameter, the average value μ _{1 of} film thickness and the standard deviation σ ₁ as characteristic values must be obtained in advance. FIG. 5 is a diagram showing the relationship between the film forming time X ₁ in the film forming process obtained from the past processing data and the average value μ ₁ of the film thickness. FIG. 6 shows the film forming time obtained from the past processing data. FIG. 4 is a diagram showing a relationship between a film forming time X ₁ in a process and a standard deviation σ ₁ of a film thickness. 5 and 6 that the average value μ ₁ and the standard deviation σ ₁ have a substantially linear relationship with the film formation time X ₁ .

By calculating a function of the regression line from this relationship, the average value / standard deviation calculating means 7 converts the film formation time X ₁ into an average value μ _{1 of} film thickness and a standard deviation σ ₁ as follows. Equations (6) and (7) were obtained.

Μ ₁ (X ₁ ) = a ₁ X ₁ + b ₁ (a ₁ and b ₁ are constants determined by the film formation rate) (6) σ ₁ (X ₁ ) = c ₁ X ₁ + d ₁ (c ₁₎ , constant d ₁ is determined by the deposition rate) ... (7)

The software in which the equations (6) and (7) were programmed was used as the average / standard deviation calculating means 7.

The standardized characteristic value calculation means 11 used software in which the above equation (3) was programmed. Since n = 1 here, the standardized characteristic value is one per wafer.

In the present embodiment, the process is managed by totalizing k = 5 standardized characteristic values.

The software obtained by programming the above equations (4) and (5) was used as the average / range calculating means 14.

The control chart display means 15 uses a display monitor of a computer, and automatically displays the average value Xbar.
And software for automatically setting the values of the range R on the display monitor.

The abnormality determining means 16 is software in which an algorithm for determining that the process is abnormal when the value of the average value Xbar exceeds ± 3 or the value of the range R exceeds +3 is used.

Further, a display monitor of a computer is used as the warning means 17, and a warning display is automatically displayed on the display monitor when a process is abnormal.

As a result of actually operating the process management system of the present invention, abnormalities in the process, such as the consumption of the aluminum target, the deterioration of the vacuum pump, and the contamination in the chamber, could be found at an early stage.

Embodiment 4 Embodiment 4 of the present invention will be described. In the fourth embodiment, a photolithography process of an electronic component is applied to the process P8. In the present embodiment, a case where a photoresist pattern is formed on a wiring metal film on a wafer of a SAW device will be described as an example. There is no difference above.

In the wafer process of the SAW device, there is a photoengraving process of forming a photoresist having a desired pattern on the surface of the metal thin film in order to process the metal thin film formed on the surface of the piezoelectric material wafer into a desired pattern. is there. The outline of the photomechanical process will be described. In the photomechanical process, first, the surface of the metal thin film is uniformly coated with a photoresist. The thickness of the photoresist is constant regardless of the type. Next, the photoresist is covered with an exposure mask on which a circuit pattern corresponding to the type is formed, and light is irradiated from above the mask to expose the photoresist. As a result, the quality of the photoresist is changed only in the portion exposed through the exposure mask. Next, only the deteriorated photoresist is removed by development. Thereby, a photoresist having a desired pattern can be formed on the metal thin film.

As described above, the frequency characteristics of the SAW device differ depending on the product type. It is widely known that one of the important factors that determines the frequency characteristic of a SAW device is the circuit wiring width (for example, “Surface Acoustic Wave Engineering” edited by IEICE, edited by Kazuo Shibayama, p.56). ). This indicates that the circuit wiring width to be formed differs depending on the product type, and that the process control of the photoengraving process must be performed in order to obtain the target line width. ing.

When the same product is processed in large numbers continuously, it is sufficient to manage the line width of the circuit wiring formed under the same exposure conditions (exposure mask type, exposure time, etc.) by the conventional method. Is changed, the exposure condition itself to be managed varies depending on the product type, and thus there is a problem in managing the photolithography process.

In this embodiment, the exposure time X ₁ [second] is selected as the processing parameter (m = 1). Since a factor to be adjusted as a factor for controlling the line width of the photoresist pattern may be selected for each type, an exposure energy or the like may be selected as a processing parameter. In this case, the same effect can be obtained. In addition, in order to unify the line width of the exposure mask pattern according to the type, the same evaluation wiring pattern 19 is provided at the central portion and the peripheral portion of the exposure mask regardless of the type. FIG. 7 is a diagram illustrating an evaluation wiring pattern for evaluating the line width of a photoresist pattern. FIG. 8 is a diagram showing a part of a cross section taken along line AA of FIG.
The evaluation wiring pattern 19 has a line portion 20 on which a photoresist is formed and a space portion 21 on which no photoresist is formed. The width of the line portion 20 and the width of the space portion 21 are both 500 nm. The evaluation wiring pattern 19 is formed at a central portion and a peripheral portion on the wafer 18. As shown in FIG. 8, the cross section of the portion where the evaluation wiring pattern 19 is formed has a wiring metal thin film 25 formed on a wafer 24 and a photoresist 20 formed on the surface thereof. The cross section of the photoresist 20 usually has a shape in which the line width 22 on the upper surface is narrow and the line width 23 on the lower surface is wide.

In the fourth embodiment, an evaluation wiring pattern 19 is formed near the center of the wafer, and as the characteristic values, the line width 22 of the upper surface of the photoresist 20 of the evaluation wiring pattern 19 is Y ₁ , The line width 23 was selected as Y ₂ (n = 2). In implementation, the characteristic values Y ₁ and Y ₂ are set to 1
One wafer was extracted from the batch (18 wafers) and measured. Since the frequency of measurement also depends on the stability of the target photolithography process, in order to more strictly control the photoresist line width at a plurality of coordinates (for example, 13 points in the plane) within the wafer plane, the characteristic value may be used. For better management, only the line width measurement of the lower surface of the photoresist may be used as the characteristic value. A scanning electron microscope (SEM) was used for measuring the photoresist line width of the wiring pattern for evaluation.

The average value / standard deviation calculation means 7 will be described. And exposure time X ₁ is a process parameter, the line width of the upper surface of the photoresist 20 of the evaluation wiring patterns 19 22
The relationship between the average values μ ₁ , μ ₂ and the standard deviations σ ₁ , σ ₂ of the line width 23 on the lower surface and the lower surface must be determined in advance. Figure 9 is a diagram showing an average value mu ₁ of the relationship between the upper surface line width of the exposure time X ₁ and the photoresist obtained from past processing data. FIG. 10 is a diagram showing the relationship between the exposure time X ₁ obtained from the past processing data and the standard deviation σ ₁ of the top line width of the photoresist. 9 and 10 that the average value μ ₁ and the standard deviation σ ₁ have a substantially linear relationship with the exposure time X ₁ . Average value μ ₂ and standard deviation σ ₂ and exposure time X
_{The same} applies to the relationship ₁ (not shown).

When the function of the regression line is determined from these relationships, the average / standard deviation calculating means 7 converts the exposure time X ₁ into the average value μ _{1 of the} line width and the standard deviation σ ₁ as follows: (8) to (11) were obtained.

Μ ₁ (X ₁ ) = a ₁ X ₁ + b ₁ (8) (a ₁ and b ₁ are constants determined by exposure energy) σ ₁ (X ₁ ) = c ₁ X ₁ + d ₁ (9) (C ₁ and d ₁ are constants determined by exposure energy) μ ₂ (X ₁ ) = a ₂ X ₁ + b ₂ (10) (a ₂ and b ₂ are constants determined by exposure energy) σ ₂ (X ₁ ) = c ₂ X ₁ + d ₂ (11) (c ₂ and d ₂ are constants determined by exposure energy)

The software obtained by programming the above equations (8) to (11) is used as the average value / standard deviation calculating means 6.

As the standardized characteristic value calculating means 12, software in which the above equation (3) was programmed was used. Here, since n = 2, there are two standardized characteristic values 10 per wafer. Hereinafter, similarly to the third embodiment, the average value / range calculation means 14 calculates the average value Xbar and the range R, the control chart display means 15 displays a control chart, and the abnormality determination means 16 determines whether or not there is a process abnormality. Judgment is made, and a warning is issued by the warning means 17 when the process is abnormal.

As a result of actually operating the process management system according to the present invention, process abnormalities such as a consumed state of an exposure lamp and an exposure energy adjustment error could be found at an early stage.

Embodiment 5 Embodiment 5 of the present invention will be described. In the fifth embodiment, the process is an electronic component etching process. In this embodiment, as an example, a case in which a wiring metal film of a SAW device is etched for forming a wiring pattern will be described. However, in the case of an etching process of another electronic component, for example, a semiconductor element, any effect is obtained. There is no difference.

In the wafer process of the SAW device, there is a process of etching a metal thin film formed on the surface of the piezoelectric material wafer after the photolithography process. An outline of the etching process will be described. In the etching step, the wafer with patterned photoresist formed in Embodiment 4 is used.
Because the photoresist masks the metal thin film,
Only the portions not covered by the photoresist are etched. After the etching, the photoresist is dissolved and peeled by a chemical treatment, so that a desired wiring pattern can be formed on the wafer surface.

In carrying out the present invention, a reactive ion etching (RIE) method was used as an etching method. Acetone and isopropyl alcohol were used for stripping and dissolving the photoresist.

As described above, the frequency characteristics of the SAW device differ depending on the product type. As described above, it is widely known that one of the important factors that determines the frequency characteristics of the SAW device is the circuit wiring width.
This indicates that the circuit wiring width to be formed differs depending on the product type, and also indicates that the process control of the etching process must be performed in order to obtain the target line width. I have.

When the same product is processed in large numbers continuously, it is sufficient to manage the line width of the circuit wiring formed under the same etching conditions (etching rate, etching time, etc.) by the conventional method. If it is changed, the etching condition itself to be managed changes depending on the product type, which causes a problem in managing the etching process.

In this embodiment, the etching time X ₁ [second] is selected as the processing parameter (m = 1). It is sufficient to select a factor to be adjusted as a factor for controlling the line width for each type, so the etching rate [% / sec]
[V] or the like may be selected as the processing parameter. Also, a plurality of factors may be selected, such as selecting both the etching time and the etching rate as processing parameters. In these cases, a similar effect can be obtained. In order to unify the line width of the wiring pattern depending on the product type, the above-described evaluation wiring pattern 1 is used.
9 was used.

FIG. 11 shows an article according to Embodiment 5 of the present invention.
FIG. 9 is a view for explaining a method of manufacturing a metal thin film.
FIG. 3 is a diagram showing a cross section after etching is performed. Form of this implementation
In this state, the wiring pattern for evaluation of the circuit wiring width is located near the center of the wafer.
The characteristic value Y ₁As this evaluation wiring
The line width 26 of the metal wiring 25 of the pattern 19 was selected (n =
1). In implementation, the characteristic value Y₁1 batch (18
(Wafer), one wafer was removed from the wafer and measured.
The frequency of measurement depends on the stability of the target etching process.
Therefore, for more strict management,
The line width of the metal wiring at a plurality of points may be used as the characteristic value. still,
For line width measurement of metal wiring of evaluation wiring pattern, use SEM
Was used.

The mean value / standard deviation calculating means 7 will be described. Setting and the deposition time X ₁ is a process parameter,
The relationship between the average value μ ₁ of the line width 26 of the metal wiring 25 of the evaluation wiring pattern 19 and the standard deviation σ ₁ , which are characteristic values, must be obtained in advance. Figure 12 is a diagram showing an average value mu ₁ of the relationship between the wiring line width and the etching time X ₁ obtained from past processing data. FIG. 13 is a diagram showing a relationship between the etching time X ₁ obtained from the past processing data and the standard deviation σ ₁ of the wiring line width. 12 and 13 that the average value μ ₁ and the standard deviation σ ₁ have a substantially linear relationship with the etching time X ₁ .

If the function of the regression line is obtained from this relationship,
In the mean, standard deviation calculation means 7, as a conversion formula for converting the etching time X ₁ to the average value mu ₁ and a standard deviation sigma ₁ line width conversion equation represented by the following formula (12) and (13) I got

Μ ₁ (X ₁ ) = a ₁ X ₁ + b ₁ (12) (a ₁ and b ₁ are constants determined by the etching rate) σ ₁ (X ₁ ) = c ₁ X ₁ + d ₁ (13) (C ₁ and d ₁ are constants determined by the etching rate)

The software in which the above equations (12) and (13) are programmed is used to calculate the average value / standard deviation calculating means 7.
And

The standardized characteristic value calculation means 11 used software in which the above equation (3) was programmed. Here, since n = 1, the standardized characteristic value 10 is one per wafer. Hereinafter, similarly to the third embodiment, the average value / range calculation means 14 calculates the average value Xbar and the range R, the control chart display means 15 displays a control chart, and the abnormality determination means 16 determines whether or not there is a process abnormality. Judgment is made, and a warning is issued by the warning means 17 when the process is abnormal.

As a result of actually operating the process management system of the present invention, process abnormalities such as fluctuations in ion acceleration voltage, fluctuations in ion energy, deterioration of the vacuum pump, and incorrect setting of conditions could be found at an early stage.

[0074]

According to the process management system of the present invention, even if the set values of the processing parameters of the process frequently change depending on the type of product to be processed, such as in the case of small-lot production of various products, whether the process is in a statistical management state. It can be managed appropriately without delay.

The article manufacturing method according to the present invention is capable of determining whether or not the process is in a statistical management state even when the set values of the processing parameters of the process frequently change depending on the type of the product to be processed, as in the case of multi-product small-lot production. Can be appropriately managed without time delay, and quality control of articles processed in the process can be appropriately performed.

Further, whether or not the film forming process of the electronic component is in the statistical management state can be appropriately managed without time delay, and the quality control of the electronic component processed in the film forming process can be appropriately performed.

Further, whether or not the photoengraving process of electronic parts is in a statistically controlled state can be appropriately managed without time delay.
Quality control of electronic components processed in the photoengraving process can be appropriately performed.

Further, whether or not the etching process of the electronic component is in the statistical management state can be appropriately managed without time delay, and the quality control of the electronic component processed in the etching process can be appropriately performed.

Further, by determining the line width of the wiring pattern by measuring the line width of the evaluation wiring pattern formed in the same pattern regardless of the product type, the line width of the wiring pattern as the characteristic value can be easily obtained. it can.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a method for manufacturing an article according to Embodiment 1 of the present invention.

FIG. 2 is a diagram showing an example of an Xbar management chart of the present invention.

FIG. 3 is a diagram showing an example of an R management chart according to the present invention.

FIG. 4 is a diagram illustrating a configuration of a process management system according to a second embodiment of the present invention.

FIG. 5 is a diagram for explaining a method of manufacturing an article according to Embodiment 3 of the present invention, and illustrates a relationship between a film forming time X ₁ and an average value μ ₁ of film thickness in a film forming process obtained from past processing data. FIG.

FIG. 6 is a diagram illustrating a method of manufacturing an article according to Embodiment 3 of the present invention, and is a diagram illustrating a relationship between a film formation time X ₁ and a standard deviation σ ₁ of a film thickness obtained from past processing data. .

FIG. 7 is a diagram illustrating a method for manufacturing an article according to a fourth embodiment of the present invention, and is a diagram illustrating an evaluation wiring pattern for evaluating a line width of a photoresist pattern.

FIG. 8 is a diagram showing a part of a cross section taken along line AA of FIG. 7;

FIG. 9 is a diagram for explaining a method of manufacturing an article according to Embodiment 4 of the present invention, and illustrates a relationship between an exposure time X ₁ obtained from past processing data and an average value μ ₁ of a top line width of a photoresist. FIG.

FIG. 10 is a diagram for explaining a method of manufacturing an article according to Embodiment 4 of the present invention, and illustrates a relationship between an exposure time X ₁ obtained from past processing data and a standard deviation σ ₁ of a top line width of a photoresist. FIG.

FIG. 11 is a diagram illustrating a method of manufacturing an article according to Embodiment 5 of the present invention, and is a diagram illustrating a cross section after etching a metal thin film in FIG. 8;

FIG. 12 is a diagram illustrating a method of manufacturing an article according to Embodiment 5 of the present invention, and is a diagram illustrating a relationship between an etching time X ₁ obtained from past processing data and an average value μ ₁ of a wiring line width. is there.

FIG. 13 is a diagram illustrating a method for manufacturing an article according to Embodiment 5 of the present invention, and is a diagram illustrating a relationship between an etching time X ₁ obtained from past processing data and a standard deviation σ ₁ of a wiring line width. is there.

[Explanation of symbols]

1 characteristic value storage means, 2 kind input means, 5 set value determination means, 7 average value / standard deviation calculation means (first calculation means), 11 standardized characteristic value calculation means (second calculation means),
14 mean / range calculating means (third calculating means), 16
Abnormality determination means, 17 warning means.

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁷ Identification code FI Theme coat ゛ (Reference) H01L 21/302 H01L 21/30 502G 21/363 21/302 Z (72) Inventor Fumihide Sato Chiyoda-ku, Tokyo 2-3-2 Marunouchi F term (reference) in Mitsubishi Electric Corporation 3C100 AA57 BB27 EE06 5B056 BB62 BB64 HH00 5F004 AA16 BA11 5F046 AA25 AA28 DA02 DD03 DD06 5F103 AA08 DD28 HH10 LL20 RR10

Claims (6)

[Claims] 1. A system for managing a process provided with a process common to a plurality of types, wherein a plurality of characteristic values after processing of each type are input for a certain process parameter in the common process. By this means, storage means for storing the characteristic values associated with each type, the average value and the standard deviation associated with each type for the characteristic value for each type based on the storage data of the storage means First calculating means for calculating, a kind input means for selecting and inputting one specific kind to be processed in the process from the plurality of kinds, and processing corresponding to a kind determining signal from the kind input means Setting value determining means for determining a set value of the parameter; a specific type corresponding to the type determination signal processed based on the set value of the processing parameter Second computing means for standardizing the processed characteristic values using the average value and standard deviation of the characteristic values associated with the specific product calculated by the first computing means; and the second computation for a plurality of processes. A third calculating means for calculating an average value and a range of the standardized characteristic value obtained by the means, and based on the average value and the range of the standardized characteristic value,
A process management system comprising: an abnormality determining unit configured to determine whether there is an abnormality in the process; and a warning unit that issues a warning when the abnormality determining unit determines that there is an abnormality. 2. A method for manufacturing an article having a process common to a plurality of types, wherein a plurality of characteristic values after processing of each type are input for a certain process parameter in the common process. Storing the characteristic values associated with each of the varieties, and an average value and a standard deviation associated with each of the varieties with respect to the characteristic values of each of the varieties based on the characteristic values associated with each of the varieties. Calculating, and inputting one specific product to be processed in the process from the plurality of products, and determining a set value of a processing parameter corresponding to the specific product. The characteristic value after processing of the specific type processed based on the set value of the processing parameter is an average of the characteristic values associated with the specific type. And standardizing using the standard deviation; calculating the average value and range of the standardized characteristic value obtained in the standardizing step for a plurality of processes; and the average value and range of the standardized characteristic value. On the basis of the,
Determining the presence or absence of an abnormality in the above process. 3. The article manufacturing method according to claim 2, wherein the step is a film forming step of the electronic component, the processing parameter is a film forming time and / or a film forming rate, and the characteristic value is a film thickness. 4. The process is a photoengraving process for electronic parts,
3. The article manufacturing method according to claim 2, wherein the processing parameter is an exposure time, and the characteristic value is a line width of the wiring pattern. 5. The method according to claim 2, wherein the step is an electronic component etching step, the processing parameter is an etching time and / or an etching rate, and the characteristic value is a line width of the wiring pattern. 6. The method for manufacturing an article according to claim 4, wherein the line width of the wiring pattern is obtained by measuring the line width of an evaluation wiring pattern formed of the same pattern regardless of the type.