JP2003332209A - Method of manufacturing semiconductor device, and method, system, and program for diagnosing production line for semiconductor device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a semiconductor device manufacturing method and a method, a system, and a program for diagnosing a production line for a semiconductor device, which are capable of reducing a non-operating time of the production line by making a diagnosis of whether the kind of the semiconductor device the manufacturing process, or the manufacturing device causes a trouble or not in the semiconductor device manufacturing line. <P>SOLUTION: It is diagnosed in an inspection process that an off-specification product is caused by a manufacturing device or a process, and the above diagnosis is made by the use of inspection data on each process in the manufacturing lines of various kinds of semiconductor devices and the actual values of the manufacturing conditions under which the manufacturing device operates in the processes. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a semiconductor device manufacturing method for manufacturing a semiconductor device, a method for diagnosing a manufacturing line for manufacturing a semiconductor device, a system therefor, and a program therefor.

[0002]

2. Description of the Related Art Semiconductor devices are manufactured through a plurality of process modules each including a plurality of element processes such as cleaning, film formation, heat treatment, impurity implantation, and lithography.

Each elemental process has a plurality of manufacturing processes, and each manufacturing process has a manufacturing apparatus group for executing the manufacturing process. Then, in each manufacturing process, one device from each manufacturing device group is used to create a device structure. Further, a plurality of steps for inspecting the workmanship (manufacturing error) in the manufacturing apparatus are included in the element process. For example, the lithographic process includes a coating process (coating a photosensitive film), an exposure process, a developing process, an etching process, and an ashing process (removing the photosensitive film). Further, a manufacturing device group such as a coating / developing device, an exposure device, an etching device, and an ashing device is used as the manufacturing device group.

Normally, in order to satisfy the electrical characteristics of a semiconductor device, in a semiconductor device manufacturing process, manufacturing errors (eg, device shape, film thickness, etc.) in each element process.
An inspection process is provided to determine whether or not is within the standard value. For example, in a lithographic process, which is a certain element process, the dimension of a transfer pattern and the overlay error between the transfer patterns of each layer are measured. Then, in each inspection step, when the measured value deviates from the standard value, the manufacturing condition of the manufacturing apparatus is adjusted, the manufacturing apparatus itself is adjusted, the process condition is changed, and the like so that the measured value is within the standard value. There are a plurality of factors that cause the inspection data to deviate from the standard value due to differences in manufacturing equipment and element processes. Therefore, an engineer usually performs an evaluation experiment to determine the factors and take countermeasures. At that time, first, the out-of-specification factors are separated into the device process and the manufacturing device, and countermeasures are taken for each factor.

[0005]

Since the manufacturing apparatus used for manufacturing a semiconductor device is expensive, it is desired to increase the throughput of the manufacturing apparatus from the economical point of view. The operating status of the manufacturing apparatus differs depending on the type of semiconductor device or the manufacturing apparatus, but in order to increase the throughput, it is important to reduce downtime such as failure, waiting, and preparation / setup time.

Usually, in the inspection process of semiconductor device manufacturing, when the inspection result is out of the standard, an evaluation experiment is manually conducted by an engineer or the like to identify the cause of the out-of-specification and a countermeasure thereof (hereinafter referred to as defect countermeasure). ing.
In a mass production line that manufactures a wide variety of devices, not only the types of devices but also the number of manufacturing apparatuses are large, so the above-mentioned countermeasures against defects are frequent, and it takes time to identify the cause itself. During this period, the manufacturing apparatus is out of order, which is a factor of decreasing the operating rate of the manufacturing apparatus.

An object of the present invention is to solve the above problems.
In a semiconductor device manufacturing line, it is possible to diagnose whether the cause of a semiconductor device failure (out-of-specification in the inspection process) is due to the type of semiconductor device, process (process), or manufacturing equipment. The present invention provides a method for manufacturing a semiconductor device, a method for diagnosing a manufacturing line, a system for the same, and a program for the same, which are capable of reducing downtime.

[0008]

In order to achieve the above object, the present invention provides a database unit in which inspection standard specification information (for example, size standard, overlay standard) of a semiconductor device is stored, and manufacturing of a manufacturing apparatus. A second database unit that stores actual values of conditions (for example, exposure energy amount and alignment correction value) and inspection data (for example, dimension inspection data, overlay inspection data) in each process of semiconductor device manufacturing are stored. The third database part for the semiconductor device, the fourth database part in which the relational expressions of the semiconductor device type and process, the manufacturing conditions and the inspection data for each manufacturing device are registered, and the diagnostic processing of the manufacturing device is performed using each database. It is composed of a diagnostic processing unit.

In the diagnostic station, a step of collecting past inspection data and actual values of manufacturing conditions for each relevant element process, and (1) net inspection data from which influence components on the inspection data due to changing manufacturing conditions are removed And (2) calculating the evaluation index of the net inspection item for each manufacturing category type combination of semiconductor device manufacturing, and (3) the process of (2) for each manufacturing category type combination of semiconductor device manufacturing. A step of calculating an integrated net evaluation index obtained by integrating a plurality of net inspection evaluation indexes calculated in step (4), the integrated net evaluation index calculated in the step (3) for each combination of manufacturing category types of semiconductor device manufacturing, and A process of determining a threshold value of the net inspection index calculated in the process of (2),
(5) A process of totaling the determination results of each evaluation index calculated in the process of (4) for each manufacturing category type of semiconductor device manufacturing, and (6) Totaling the determination results of each manufacturing category type of semiconductor device manufacturing A diagnostic processing step having a process of extracting a problematic part from the value, and the diagnostic processing unit sequentially performs the processes of (1) to (6), thereby performing diagnostic processing of the manufacturing apparatus to perform diagnostics. The problem location is extracted based on the result.

As a result, the time required for identifying the nonstandard factor in the inspection process can be shortened, a quick countermeasure can be taken against the defect, the downtime of the manufacturing apparatus due to the defect can be reduced, and the yield of semiconductor devices can be improved at an early stage. .

Further, the present invention resides in the predetermined manufacturing process in a manufacturing line for manufacturing a semiconductor device through a plurality of process modules including a plurality of element processes having a plurality of manufacturing processes including a manufacturing apparatus group. Based on the inspection result obtained by inspecting the semiconductor substrates manufactured by each of the plurality of manufacturing apparatuses in the manufacturing apparatus group for each factor (process) and the manufacturing condition set for each of the plurality of manufacturing apparatuses. The calculated evaluation index, for each predefined manufacturing category combination (for example, a combination of product type, process, and manufacturing equipment), is aggregated for a plurality of semiconductor substrates,
A diagnostic processing unit that extracts a manufacturing category indicating a defect based on the aggregated evaluation index and outputs information on a manufacturing category indicating a defect is provided, and the information on the manufacturing category indicating a defect output from the diagnostic processing unit is displayed. A method for manufacturing a semiconductor device is characterized in that a semiconductor device is manufactured by taking measures based on the above.

Further, according to the present invention, in a manufacturing line for manufacturing a semiconductor device through a plurality of process modules including a plurality of element processes having a plurality of manufacturing processes in which a group of manufacturing apparatuses exists, the present invention exists in each step for each corresponding element process. Manufacturing when a plurality of semiconductor substrates manufactured by the manufacturing apparatus are collected by the inspection apparatus and actual values of past inspection data obtained by the inspection apparatus are collected and accumulated in a database, and the plurality of semiconductor substrates are manufactured by the manufacturing apparatus. An accumulation step of collecting actual values of conditions and accumulating them in a database, and in the diagnostic processing unit, for each combination of manufacturing category type when the semiconductor substrate is manufactured (for example, combination of product type, process, and manufacturing apparatus), From the actual values of the past inspection data accumulated in the accumulation step, accumulated in the accumulation step A net inspection data calculating step of calculating net inspection data from which the influence component on the past inspection data due to the change of the manufacturing condition in the actual value of the manufacturing condition is removed, and the manufacturing calculated in the net inspection data calculating step A net inspection evaluation index calculation process for calculating a net inspection evaluation index based on a plurality of net inspection data for each category type combination, and a net inspection evaluation index for each manufacturing category type combination calculated in the net inspection evaluation index calculation process The integrated net evaluation index calculation step of calculating an integrated integrated net evaluation index, the net inspection index calculated in the net inspection evaluation index calculation step and the integrated net evaluation index calculated in the integrated net evaluation index calculation step Index judgment totals that judge each combination of category types and total the judgment results And a problem location extraction process of extracting information on a problem location in the production line based on the aggregated value of the determination results for each production category type in the index determination aggregation process, and performing the diagnostic processing of the production line. And a diagnostic processing step, wherein a semiconductor device is manufactured by taking countermeasures based on the information of the problematic point in the manufacturing line extracted in the problematic point extracting process of the diagnostic processing step. Is.

[0013]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of a method for diagnosing a semiconductor device manufacturing line, a system therefor and a program therefor, and a method for manufacturing a semiconductor device according to the present invention will be described with reference to the drawings.

FIG. 1 schematically shows a process of manufacturing a certain type of semiconductor device. A semiconductor device of a certain type includes process modules (corresponding to layers) 1 to 4 ...
Is manufactured through multiple processes. Element process 1, 2, 3, 4,
… As impurity injection, film formation, lithography, cleaning,
Examples include heat treatment. As a matter of course, if the type of semiconductor device changes, the steps that go through the elemental processes for each process module may change. Also, if the type of semiconductor device changes, the manufacturing process in the element process described below changes in most cases.

FIG. 2 schematically shows a manufacturing process in a certain element process. In a certain elemental process, a plurality of manufacturing processes ... i_3 to i_5 ... Exist, and manufacturing device groups realizing each manufacturing process ..., i_3, i_4, i_5, ..., Mc (51c), Md (51d), Me.
(51e), ... Are present. Then, in each manufacturing process ..., i_3, i_4, i_5, ..., each manufacturing device group ..., Mc (51c), Md (51d), Me (51).
e), ..., One of them is used to make up the device structure. Further, a certain element process includes a plurality of inspection steps for inspecting the workmanship (manufacturing error) in the manufacturing apparatus. For example, when the element process is a lithography process, a plurality of manufacturing processes ... i_3 to i_5
Examples of the processes include a coating process (coating a photosensitive film), an exposure process, a developing process, an etching process, and an ashing process (removing the photosensitive film). Also,
Manufacturing apparatus group ... Mc to Me ... (51) includes coating / developing apparatus, exposure apparatus, etching apparatus, ashing apparatus, CMP (Chemical Mechanical) corresponding to the above manufacturing process.
A manufacturing device group such as an alPolishing device is used.

In the embodiment of the present invention, for example, the manufacturing apparatus group Mc (5) in the lithography process 3 which is one of the element processes for manufacturing a semiconductor device (semiconductor substrate).
1c), a method of diagnosing overlay error between transfer patterns in the exposure apparatus, a system thereof, a method of manufacturing a semiconductor device, and a manufacturing line thereof will be described.

FIG. 3 shows the construction of a reduction projection exposure apparatus mainly used in the exposure process i_3. In the reduction projection exposure apparatus, for example, a circuit pattern drawn on a mask 42 illuminated by ultraviolet light such as excimer laser light by an illumination system 41 is passed through a reduction lens 43 to a wafer (semiconductor substrate) 4
The image is transferred onto the wafer 44 by successively performing reduction exposure on the surface of the wafer 4 by 1 to several chips. Further, for example, by exchanging the mask 42 corresponding to the process module, the circuit pattern for each layer is formed on the wafer 44. At this time, in order to align the pattern formed on the wafer 44 and the reduced-projection transfer pattern, the position of the base pattern on the wafer is detected by the alignment optical system 46, and the position is adjusted to that position. The wafer stage 45 is driven to perform sequential exposure processing.

FIG. 4 shows that the element process is a lithography process, and the exposure process i_ in the lithography process.
3 shows an example of an overlay error (past inspection item and manufacturing condition item) that occurs in the reduction projection exposure apparatus which is No. 3. In the exposure apparatus, alignment is performed during the exposure process, but an error as shown in FIG. 5 occurs due to an alignment detection error due to a change in the quality of the underlying pattern on the wafer and the stability of the exposure apparatus itself.
Further, the exposure apparatus is provided with a function of compensating for the error shown in FIG. 5, so that the overlay error can be adjusted. For example, when the offset error is 100 nm in the X direction, the offset correction value of the exposure apparatus is 100 nm.
By inputting, the -100 nm wafer stage is forcibly driven in the X direction to perform the exposure process, and the overlay error is corrected. In the case of the reduction projection exposure apparatus, the manufacturing condition items as the exposure apparatus include, for example, exposure energy amount and alignment correction value (offset corresponding to each error shown in FIG. 5, wafer scale (expansion and contraction), wafer rotation, wafer orthogonality). Degree, shot rotation, shot magnification correction value) and the like. Examples of past inspection items (inspection data) for wafers exposed by the exposure apparatus include dimension inspection data and overlay inspection data.

When the element process is a lithography process and the etching apparatus is an etching process in the lithography process, plasma generation power, introduced etching gas pressure and the like can be considered as manufacturing condition items. As past inspection items (inspection data) for wafers etched by the etching device,
The average wafer size as the etching size, the variation within the wafer surface, and the variation within the shot are considered.

FIG. 5 is a block diagram showing the overall construction of the first embodiment of the diagnostic system for the semiconductor device manufacturing line according to the present invention. In the figure, manufacturing equipment groups (..., Mc (51c), Md (51d), Me (51
e), ...) 51 is composed of at least one manufacturing apparatus 52. The manufacturing apparatus group 51 includes a manufacturing apparatus group 51 having a plurality of different functions corresponding to the element processes 1 to 4, .... The manufacturing device group 51
Is connected to a diagnostic station 55 having a diagnostic processing unit 56, an input / output interface 57, and a database unit 58. The manufacturing apparatus group 51 is connected to a manufacturing management system 59 that manages the entire semiconductor device manufacturing line. It should be noted that the type, process, and wafer name of the semiconductor device of the wafer at the time of processing can be input by, for example, reading the number formed on the wafer and transmitted from the manufacturing management system 59 via the network 61. It is possible.

The inspection device group 53 comprises at least one inspection device 54, and a diagnostic station 55.
And the manufacturing control system 59. Also,
The inspection device group 53 is composed of a plurality of inspection device groups having different functions according to the inspection function. For example, as shown in FIG. 2, the inspection steps 21c and 21 in a certain element process are performed.
will be used in e.

The database section 58 includes inspection standard specification information (for example, dimensional standard, etc.) for each type of semiconductor device and each process.
A semiconductor device specification database 581 for accumulating overlay standards) and a manufacturing condition result database 582 for accumulating actual values of manufacturing condition items for semiconductor devices (wafers: semiconductor substrates) manufactured for each product type, process, and manufacturing apparatus. And an inspection device 54 for a semiconductor device (wafer: semiconductor substrate) obtained for each product type, process, and manufacturing device.
Inspection result database 583 in which actual values of past inspection data that have been inspected are accumulated, and manufacturing condition conversion formulas in which manufacturing condition conversion formulas for semiconductor devices (wafers) manufactured for each product type, process, and manufacturing apparatus are registered. And the expression database 584. When the manufacturing apparatus is an exposure apparatus, the manufacturing condition items accumulated in the manufacturing condition result database 582 as actual values include, for example, the exposure energy amount and the alignment correction value, and the inspection result database 5
Examples of past inspection data accumulated as actual values in 83 include dimension inspection data and overlay inspection data. In response to the inquiry from the diagnostic processing unit 56, the database unit 58 searches for the corresponding data via the input / output interface 57, and the answer data is transmitted to the diagnostic processing unit 56.

The diagnostic processing unit 56 is a diagnostic program storage unit.
Diagnostic program stored in 60 (standard value such as initial value
Inspection data peculiar to manufacturing equipment under manufacturing conditions (Net item
Eye value calculation program (step 102 shown in FIG. 6)
Program to be executed) and the network for each manufacturing category combination
Program for calculating item evaluation value (Step 1 shown in FIG. 6
03), each manufacturing category combination
Integrated net evaluation value calculation program (see the
Program for executing the setup 104), manufacturing category group
Threshold value judgment program for each evaluation value for each combination (shown in Fig. 6
Program for executing step 105), manufacturing category
A program that aggregates each evaluation value for each gori (see
Program that executes step 106), problem area (manufactured
Extraction (Picking category or Manufacturing category combination)
Output) Program (Execute step 107 shown in FIG. 6)
Acquired from the database unit 58 according to the program))
Diagnosis and diagnosis for each element process based on the data
The disconnection result is output to the diagnostic screen via the input / output interface 57.
Station 55 itself, desired manufacturing equipment 52 or manufacturing pipe
Output to the processing system 58 or the like.

The manufacturing management system 59 uses a wafer (semiconductor substrate) as a manufacturing apparatus in a process for manufacturing a semiconductor device.
The manufacturing process is instructed, and the manufacturing process is ordered based on information such as the operating rate of each manufacturing device 52 in the manufacturing device group 51 that performs the manufacturing process of the semiconductor device (semiconductor substrate). Also,
Regarding the inquiry of the wafer number from the manufacturing apparatus group 51 and the inspection apparatus group 53, the type of semiconductor device (wafer),
Answer the process. Further, the manufacturing control device 59 can also output the information of the problem portion, which is the diagnosis result diagnosed by the diagnosis station 55, to a display device or the like, and the diagnosis result can be displayed to the manufacturing device or the manufacturing process of the problem portion. It is also possible to feed back information on a certain problem area and take measures to eliminate the problem. Naturally, manufacturing conditions (in the case of an exposure apparatus, exposure energy, alignment correction values (offset, wafer scale (expansion and contraction), wafer rotation,
Wafer orthogonality, shot rotation, and shot magnification correction value) are considered, and in the case of an etching apparatus, plasma generation power, introduced etching gas pressure, etc. are considered. ), If it is possible to eliminate the problem, it is possible to automatically take measures.

Next, an embodiment of the method for diagnosing the manufacturing apparatus according to the present invention will be described with reference to the flowchart shown in FIG. First, in step 101, the diagnostic processing unit 56 in the diagnostic station 55 relates to an element process (for example, a lithography process) to be diagnosed,
The past inspection item data and manufacturing condition item data for each semiconductor device manufacturing category type (product type, process, manufacturing apparatus) are searched from the databases 583 and 582 which store inspection data and manufacturing condition data. At this time, processing for matching the wafer ID with the manufacturing category information of the manufacturing condition data and the inspection data is performed. Examples of manufacturing category types in manufacturing semiconductor devices include product types, processes, and manufacturing equipment. In the case of superposition in the exposure process, a combination of the underlayer exposure device and the exposure device is also listed as a manufacturing category type. That is, the same manufacturing category (product type, process, manufacturing device) corresponds to the manufacturing condition data and the inspection data.

In the case where the element process is a lithography process and the overlay exposure process therein, the past inspection item data (past overlay inspection data) inspected by the inspection device 54 is shown in FIG. Inspection items such as offset (X and Y directions), wafer scale (X and Y directions), wafer rotation, wafer orthogonality, shot rotation, and shot magnification error are included. Further, in the case of the exposure apparatus used for the overlay exposure process, the manufacturing condition item data that is set or controlled includes offsets (X and Y directions), wafer scales (X and Y directions), and wafer rotations as alignment correction values. Condition items such as wafer orthogonality, shot rotation, and shot magnification correction value are included.

Next, the diagnostic processing section 56 determines in step 102.
In the inspection item data retrieved in step 101, the inspection item data obtained by removing the change amount of the inspection item data due to the change of the manufacturing condition (hereinafter,
Net item value) is calculated. The calculated net item value corresponds to inspection item data inspected for a semiconductor device (wafer) in the case of being manufactured under the same manufacturing condition (for example, standard manufacturing condition such as initial value) in a certain manufacturing apparatus. At this time, the relational expression between each manufacturing condition item and each inspection item data (hereinafter referred to as a manufacturing condition conversion formula) is registered in the database 584 in advance, and is sequentially called from the database 584 for use.

In the case of overlay exposure, the net item value for each error item (the same manufacturing condition (for example, exposure energy amount,
Shows the ability when exposed with the alignment correction value). ) Is calculated by the following equation (1).

Net item value = error item value + correction item value (1) The error item value is inspection data. The correction item value is a value for removing the amount of change in the inspection data due to the intentional change of the manufacturing conditions in the exposure apparatus. When the exposure energy amount is used as the manufacturing condition, a variable or a coefficient is required as the correction item value.

Next, the diagnostic processing section 56 determines in step 103.
At, the net item value calculated in step 102 is used to calculate a net item evaluation value for each manufacturing category type combination. However, if the number of data for each manufacturing category type combination is less than or equal to the threshold set number, that is, if the number is small, the reliability is poor and the net item evaluation value is not calculated. In the case of superposition, for example, the standard deviation of each net item value is used as the net item evaluation value of each error item.

Next, the diagnostic processing unit 56 determines in step 104.
In, the integrated net evaluation value is calculated for each manufacturing category combination using the net item value calculated in step 102. The integrated net evaluation value is the total net inspection data variation amount, which is the effect of each net item value added. In the case of superposition, the integrated net evaluation value is calculated by the following equation (2). Next, in step 105, the diagnostic processing unit 56 determines the net item evaluation value for each manufacturing category type combination calculated in step 103, the integrated net evaluation value calculated in step 104, and the threshold value for each item. Judgment processing is performed by comparison with. In the case of superposition, the net item evaluation value and the integrated net evaluation value are the standard deviation of the superposition error.

Next, the diagnostic processing section 56 determines in step 106.
In, the NG determination occurrence rate for each manufacturing category item is calculated for each manufacturing category type. In addition, the analysis of variance of the NG occurrence frequency for each manufacturing category item is performed to test the significant difference in the NG occurrence frequency between the manufacturing category items.

A method for determining a significant difference by analysis of variance will be described below. The case where the manufacturing category items i are _{1 to m and} the data j corresponding to each manufacturing category item is n _{1 to} n _m will be described. The data in each manufacturing category item is represented by Xij. First, the variation SA between manufacturing category items is calculated by the equation (3).

Next, the variation SE within the manufacturing category item is calculated by the equation (4). Further, the inter-manufacturing category variation degree of freedom KA and the intra-manufacturing category variation degree of freedom KE are obtained by the following equations (5) and (6). Then, using SA, SE, KA, and KE, the following (7)
The F value is calculated as in the formula. F from KA, KE, F value using the probability density function of F distribution
The distribution probability is calculated, and it is determined whether there is a significant difference between manufacturing category items from the integrated table. Generally, 0.0
When it is 5 or less, it is determined that there is a significant difference.

In the case of superposition, the NG determination occurrence rate is calculated for each product type, process, exposure apparatus, and combination of exposure apparatus and underlayer exposure apparatus.

Next, the diagnostic processing section 56 determines in step 107.
In the order of descending NG determination occurrence rate, the manufacturing category items (problem points) in question are prioritized, and the diagnostic station 55, the manufacturing device group 51, or the display device of the manufacturing management system 59 is output. Output to the device.

FIG. 7 shows steps 103 and 104 of FIG.
Example of calculation result of net item evaluation value and integrated net evaluation value for each manufacturing category type combination in FIG. 6 and threshold judgment result of net item evaluation value and integrated net evaluation value for each manufacturing category type combination in step 105 of FIG. FIG. In the figure, there are three manufacturing category types A to C, each manufacturing category type A has two items, B has three items, and C has three items. The net item evaluation value of 1 for each manufacturing category type combination, .., integrated net evaluation, and the determination result for each are displayed.

For example, the manufacturing category A is a semiconductor device type, the manufacturing category B is a step of the same element process shown in FIG. 1, and the manufacturing category C is a manufacturing apparatus in a certain manufacturing apparatus group used in the same element process shown in FIG. , The product types are A1 and A2, and the process is B1, B2, B
The three manufacturing processes of C3, C3, C3, and C3 are targeted. Then, the diagnosis processing unit 56
The database unit 58 is processed for each combination of manufacturing apparatuses.
.. (offset (X, Y direction), wafer scale (X, Y direction), wafer rotation, wafer orthogonality from the semiconductor device inspection data and manufacturing condition data stored in , Shot rotation, shot magnification error evaluation value) and integrated net evaluation value are calculated, and a threshold value determination is performed for each evaluation value to determine whether it is OK or NG.

FIG. 8 shows steps 106 and 10 of FIG.
7 is a diagram showing an example of results of totaling according to manufacturing categories in FIG. The NG occurrence rate based on the judgment of the integrated net evaluation value for the manufacturing category A1 is 2/3 = 0.67, and the NG occurrence rate based on the judgment of the integrated net evaluation value for the manufacturing category A2 is 1/3 = 0.33. It is judged that there is a significant difference. The NG occurrence rate based on the determination of the integrated net evaluation value for the manufacturing category B1 is 2/2 = 1, and the manufacturing category B is
The NG occurrence rate based on the judgment of the integrated net evaluation value for 2 is 0/2 = 0, and the NG occurrence rate based on the judgment of the integrated net evaluation value for the manufacturing category B3 is 1/2 = 0.5, which means that there is a significant difference. To be judged. The NG occurrence rate based on the judgment of the integrated net evaluation value for the manufacturing category C1 is 1/2 =
0.5, NG occurrence rate based on determination of integrated net evaluation value for manufacturing category C2 is 1/2 = 0.5, N based on determination of integrated net evaluation value for manufacturing category C3
The G occurrence rate is 1/2 = 0.5, and it is judged that there is no significant difference.

As described above, item 1 of manufacturing category type A
It can be seen that the NG occurrence rate, which is the threshold determination result of the integrated net evaluation value in item 1 of manufacturing category type B, is the highest and the priority is high. In addition, as a factor of this, the NG occurrence rate of the net item evaluation value of 1 is large. From this, it can be seen that a defect has occurred with respect to inspection item 1 in item 1 of manufacturing category types A and B.

For example, manufacturing category A is a semiconductor device type, manufacturing category B is a step of the same element process shown in FIG. 1, manufacturing category C is a manufacturing apparatus in a certain manufacturing apparatus group used in the same element process shown in FIG. In such a case, defects are concentrated in the product type A1 and the process B1. It can be determined that this is a defect caused by the type and process (process) that are not dependent on the manufacturing apparatus represented by C1 to C3.

Further, as a cause of this, when the inspection item 1 is an offset in the overlay exposure, it can be seen that a problem occurs in the setting of the offset.

Of course, between the manufacturing categories C1 to C3, N
If there is a significant difference due to a change in the G occurrence rate,
This means that the manufacturing apparatus of manufacturing category C having a high NG occurrence rate has a problem.

FIG. 9 shows steps 103 and 10 of FIG.
6 is a diagram showing a method of calculating a net item value and an integrated net evaluation value in FIG. The net item values 1, 2, ... Are manufacturing condition items 1, 2 ,.
Calculated from ... Then, the integrated net evaluation value is calculated from the evaluation index of each net item value 1, 2, ....

FIG. 10 is a block diagram showing a second embodiment of the diagnostic system for the semiconductor device manufacturing line according to the present invention. The parts equivalent to those in FIG. 5 are designated by reference numerals. In this embodiment, a diagnostic station 55 including a database unit 58, a diagnostic processing unit 56, and a diagnostic program storage unit 60 is a manufacturing management system 59.
It is built in and each processing is carried out.

In the above embodiments, the case where the present invention is applied to the superposition of the exposure processing in the semiconductor device manufacturing is specifically shown, but the present invention is particularly limited to the superposition of the exposure processing in the semiconductor device manufacturing. However, the present invention can be similarly applied to other element processes for manufacturing other devices.

[0047]

According to the present invention, in the production lines of various kinds of semiconductor devices, the cause of the out-of-specification caused by the inspection is
As it has a diagnostic processing function that can be classified as one of semiconductor device type, process (process), and manufacturing equipment, the time required to identify out-of-specification factors can be shortened, and quick countermeasures can be taken. The downtime of the manufacturing apparatus or the like is reduced, and the yield of semiconductor devices can be improved at an early stage.

[Brief description of drawings]

FIG. 1 is a diagram showing a semiconductor device manufacturing line and a manufacturing method thereof according to the present invention.

FIG. 2 is a diagram showing a manufacturing process, a manufacturing device group, and an inspection device in each of the element processes shown in FIG.

FIG. 3 is a diagram showing a configuration of a reduction type projection exposure apparatus according to the present invention.

FIG. 4 is a diagram showing an overlay error in the reduction type projection exposure apparatus according to the present invention.

FIG. 5 is a block diagram showing a first embodiment of a diagnostic system in a semiconductor device manufacturing line according to the present invention.

FIG. 6 is a flowchart showing a diagnostic method in a semiconductor device manufacturing line according to the present invention.

FIG. 7 is a diagram showing an embodiment of an output screen of an evaluation index value and its determination result according to the present invention.

FIG. 8 is a diagram showing an embodiment of a diagnostic result output screen according to the present invention.

FIG. 9 is a diagram showing a method of calculating a net item value and an integrated net evaluation value according to the present invention.

FIG. 10 is a block diagram showing a second embodiment of a diagnostic system in a semiconductor device manufacturing line according to the present invention.

[Explanation of symbols]

41 ... Illumination system, 42 ... Mask, 43 ... Reduction lens, 44 ...
Wafer (semiconductor substrate), 45 ... Wafer stage, 46 ... Alignment optical system, 51 ... Manufacturing device group, 52 ... Manufacturing device,
53 ... Inspection device group, 54 ... Inspection device, 55 ... Diagnosis station, 56 ... Diagnosis processing unit, 57 ... Input / output interface, 58 ... Database unit, 581 ... Semiconductor specification database, 582 ... Manufacturing condition result database, 583 ...
Inspection record database, 584 ... Manufacturing condition conversion type database, 59 ... Manufacturing management system, 61 ... Network.

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Kazuyuki             5-20-1 Kamimizuhonmachi, Kodaira-shi, Tokyo Stock             Ceremony Company within Hitachi Semiconductor Group (72) Inventor Yoshiyuki Miyamoto             5-20-1 Kamimizuhonmachi, Kodaira-shi, Tokyo Stock             Ceremony Company within Hitachi Semiconductor Group (72) Inventor Tetsuji Yokouchi             5-20-1 Kamimizuhonmachi, Kodaira-shi, Tokyo Stock             Ceremony Company within Hitachi Semiconductor Group F term (reference) 3C100 AA57 AA58 BB03 BB27 EE06                 5F046 AA28 DA30 DD06                 5H223 AA05 DD09 EE06 FF06

Claims (12)

[Claims] 1. A manufacturing line for manufacturing a semiconductor device through a plurality of process modules including a plurality of elemental processes having a plurality of manufacturing processes in which a manufacturing device group exists, in a manufacturing device group existing in the predetermined manufacturing process. An evaluation index calculated based on the inspection result inspected for each factor for the semiconductor substrate manufactured by each of the plurality of manufacturing apparatuses and the manufacturing conditions set for each of the plurality of manufacturing apparatuses,
A diagnostic processing unit that aggregates a plurality of semiconductor substrates for each predefined manufacturing category combination, extracts a manufacturing category indicating a defect based on the aggregated evaluation index, and outputs information on the manufacturing category indicating a defect And manufacturing a semiconductor device by taking countermeasures based on information of a manufacturing category indicating a defect output from the diagnostic processing unit. 2. In a manufacturing line for manufacturing a semiconductor device through a plurality of process modules each including a plurality of element processes having a plurality of manufacturing processes including a manufacturing apparatus group, a manufacturing apparatus existing in each step of the corresponding element process is used. Actual values of past inspection data obtained by inspecting a plurality of manufactured semiconductor substrates with an inspection device are collected and accumulated in a database, and actual values of manufacturing conditions when a plurality of semiconductor substrates are manufactured by the manufacturing device. An accumulation step of collecting and accumulating in a database, and the diagnosis processing unit, from the actual value of the past inspection data accumulated in the accumulation step, for each combination of manufacturing category types when the semiconductor substrate is manufactured, The above-mentioned past due to changing the manufacturing condition in the actual value of the manufacturing condition accumulated in step Net inspection data calculation process for calculating the net inspection data from which the influence component on the inspection data is removed, and the net inspection evaluation based on the plurality of net inspection data for each manufacturing category type combination calculated in the net inspection data calculation process. A net inspection evaluation index calculation step of calculating an index, and an integrated net evaluation index calculation step of calculating an integrated net evaluation index that integrates the net inspection evaluation indexes for each manufacturing category type combination calculated in the net inspection evaluation index calculation step An index for judging the net inspection index calculated in the net inspection evaluation index calculation step and the integrated net evaluation index calculated in the integrated net evaluation index calculation step for each manufacturing category type combination, and totaling the judgment results Aggregated value of the judgment result for each manufacturing category type in the judgment aggregation process and the index judgment aggregation process And a diagnostic processing step of performing diagnostic processing on the manufacturing line by extracting information on a defective location in the manufacturing line based on the above, and extracting in the problem location extracting step of the diagnostic processing step. A method of manufacturing a semiconductor device, wherein a semiconductor device is manufactured by taking countermeasures based on information on a problematic part in the manufacturing line. 3. The inspection data obtained by changing the condition value of the manufacturing condition by using a relational expression of the condition value of the manufacturing condition and the inspection data registered in advance in the net inspection data calculation process of the diagnostic processing step. 3. The method for manufacturing a semiconductor device according to claim 2, wherein the net inspection data is calculated by correcting the fluctuation amount of. 4. A determination process of performing threshold value determination on an integrated net evaluation index and a net inspection index in the index determination aggregation process of the diagnostic processing step, and a determination result of each evaluation index determined in the determination process is manufactured. 3. The method of manufacturing a semiconductor device according to claim 2, further comprising a totaling process of totaling for each category type. 5. The method of manufacturing a semiconductor device according to claim 2, wherein in the process of calculating the net inspection evaluation index in the diagnostic processing step, the degree of variation of the net inspection data is used as the evaluation index. 6. The semiconductor device according to claim 2, wherein the integrated net evaluation index is calculated by a weighted sum of squares of a plurality of net inspection evaluation indexes in the integrated net evaluation index calculation step of the diagnostic processing step. Production method. 7. A variance analysis of aggregated values of determination results for each manufacturing category type of semiconductor device manufacturing is performed in the problem location extraction process of the diagnostic processing step to determine a significant difference between items of the same manufacturing category type. The method for manufacturing a semiconductor device according to claim 2, wherein 8. The semiconductor device according to claim 2, wherein, in the problem location extraction process of the diagnostic processing step, the aggregated values of the determination results for each semiconductor device manufacturing category are sorted and output according to the frequency of occurrence. Device manufacturing method. 9. The method of manufacturing a semiconductor device according to claim 1, wherein the manufacturing category type includes a type of semiconductor device, a process, a manufacturing apparatus, and a manufacturing apparatus in which an underlayer is formed. 10. In a manufacturing line for manufacturing a semiconductor device through a plurality of process modules each including a plurality of element processes having a plurality of manufacturing processes including a manufacturing apparatus group, a manufacturing apparatus existing in a step for each corresponding element process is used. A first database unit that stores actual values of past inspection data obtained by inspecting a plurality of manufactured semiconductor substrates with an inspection apparatus, and manufacturing conditions of a plurality of semiconductor substrates manufactured by the manufacturing apparatus. A second database unit that stores actual values, and for each manufacturing category type combination when the semiconductor substrate is manufactured, from the actual value of the past inspection data accumulated in the first database unit, the second To the past inspection data by changing the manufacturing condition in the actual value of the manufacturing condition accumulated in the database part A net inspection data calculation process for calculating the net inspection data from which the influential components are removed, and a net inspection evaluation index is calculated based on a plurality of net inspection data for each manufacturing category type combination calculated in the net inspection data calculation process. A net inspection evaluation index calculating step, an integrated net evaluation index calculating step of calculating an integrated net evaluation index obtained by integrating the net inspection evaluation indexes for each manufacturing category type combination calculated in the net inspection evaluation index calculating step, and the net inspection An index determination aggregation process of determining the net inspection index calculated in the evaluation index calculation process and the integrated net evaluation index calculated in the integrated net evaluation index calculation process for each of the manufacturing category type combinations, and totaling the determination results; , Based on the aggregate value of the determination result for each manufacturing category type in the index determination and aggregation process, Diagnostic system of the production line, characterized in that a diagnostic unit for diagnosing process of the production line and a defect portion extraction step of extracting information of problem areas in concrete line. 11. A method for diagnosing in a diagnostic processing unit a manufacturing line for manufacturing a semiconductor device through a plurality of process modules including a plurality of element processes having a plurality of manufacturing processes in which a manufacturing apparatus group exists, wherein the predetermined manufacturing is performed. Calculated based on the inspection result obtained by inspecting the semiconductor substrate manufactured by each of the plurality of manufacturing apparatuses in the manufacturing apparatus group existing in the process by factor and the manufacturing condition set for each of the plurality of manufacturing apparatuses. The evaluation index
A characteristic is that a plurality of semiconductor substrates are aggregated for each predefined manufacturing category combination, a manufacturing category indicating a defect is extracted based on the aggregated evaluation index, and information of the manufacturing category indicating a defect is output. Diagnostic method for manufacturing line. 12. A program for diagnosing in a diagnostic processing unit a manufacturing line for manufacturing a semiconductor device through a plurality of process modules including a plurality of element processes having a plurality of manufacturing processes including a manufacturing apparatus group, wherein the semiconductor substrate is From the actual value of accumulated past inspection data for each manufacturing category type combination at the time of production, remove the influence component on the past inspection data due to changing the manufacturing condition in the accumulated actual value of manufacturing condition And a net inspection evaluation index for calculating a net inspection evaluation index based on a plurality of net inspection data for each manufacturing category type combination calculated in the net inspection data calculation process. Calculation process and manufacturing categorization calculated in the net inspection evaluation index calculation process (1) In the integrated net evaluation index calculation process for calculating the integrated net evaluation index that integrates the net inspection evaluation indices for each type combination, and the net inspection index and the integrated net evaluation index calculation process calculated in the net inspection evaluation index calculation process. Based on the index determination aggregation process for determining the calculated integrated net evaluation index for each manufacturing category type combination and totalizing the determination results, and the aggregated value of the determination results for each manufacturing category type in the index determination aggregation process And a problem location extraction process for extracting information on a problem location in the production line.