JP4231719B2 - Production scheduling method and production scheduling system in semiconductor device manufacturing - Google Patents

Description

【００２５】
また、同一品種でも露光装置によって処理時間が異なるため、投入ロットに対する負荷が異なる。そこで、各露光装置の各ファイン工程の処理時間の和を全ＴＡＴで割った値を投入ロットに当たり負荷の係数ａ(m)とする。投入後の各露光装置の負荷B(m)は（４）式にて表される。
【００２６】
B(m)=A(m)+a(m)・X(m) ・・・（４）
投入後の露光装置の負荷B(m)を平準化するように、各露光装置の投入ロット数はX(m)を決定すれば良い。ただし、このとき（３）式の制約条件を満たす必要がある。このような最適化問題の解法として、オペレーションリサーチ手法の動的計算法がある。本手法では、（３）式の制約条件の元で評価関数g(B(m))を最大化するX(m)を求めることができる。
【００２７】
次に、ステップ４０３において、ステップ４０１にて算出した投入負荷の配分結果における評価指標を算出する。評価指標として、各露光装置の負荷の標準偏差や平均を用いると良い。ステップ４０４では、組合せＮｏが総組合せ数より小さいの場合は、組合せＮｏを更新し、ステップ４０２に戻る。組合せＮｏが総組合せ数と等しい場合は、次ステップに進む。ステップ４０５では、ステップ４０３にて算出した組合せＮｏ毎の評価指標のうち、最大（または最小、評価指標の定義方法により異なる）となる組合せＮｏを抽出し、その際の割付装置リストと各装置への総投入ロット数を読み込み、最適割付装置組合せとする。
【００２８】
図１５は図２のステップ２０９における日別生産計画の算出方法を示すフローチャートである。まず、ステップ５０１において、ステップ５０２〜５０７までを割付処理の対象日数分の日別装置割付処理を繰り返す処理を定義する。ステップ５０２では、ステップ５０３〜５０６までを割付処理対象となる投入品種数分の日別装置割付処理を繰り返す処理を定義する。ステップ５０３において、ある投入日、品種の投入ロットが存在する場合、該当投入ロットが割付可能である露光装置かつ該当品種・露光装置の投入ロット総数が図２のステップ２０８で算出した総投入ロット数より小さい露光装置のうち、その日時点において装置負荷が最小となる露光装置を算出する。次に、ステップ５０４において、ステップ５０３にて算出した装置負荷が最小である露光装置を該当ロットの使用露光装置として割り付ける。ステップ５０５において、ステップ５０４にて割り付けた投入ロットの負荷を該当する露光装置に割り付ける。負荷の算出方法は、図２のステップ２０５の初期負荷の算出と同じように、各ファイン工程の処理時間の和を全ＴＡＴで割った値に投入ロット数を掛けた値を該当する露光装置の負荷に追加する。このとき、露光装置に着工されている品種の数により段取り時間が異なることもあり、投入負荷に露光装置毎の投入品種数の関数で表される段取り補正係数を加味する事もある。また、実際にはロットをまとめてラインに投入するため、まとめロット数を考慮したバッチ単位に露光装置の割付処理を行う。例えば、品種Ａのある日の投入ロット数＝１２、まとめロット数＝４の場合、バッチ数は３となるためステップ５０３からステップ５０５までの計算を３回に分けて行う。ステップ５０６では、品種Ｎｏが投入品種数より小さいの場合は、品種Ｎｏを更新し、ステップ５０３に戻る。品種Ｎｏが投入品種数と等しい場合は、次ステップに進む。ステップ５０７では、投入日が割付処理の終了日より小さい場合は、投入日を更新資、ステップ５０２に戻る。投入日が割付処理の終了日と等しい場合は処理を終了する。
【００２９】
図１６は本発明の第１の実施形態の全体構成を示すブロック図である。同図において、露光装置群１は少なくても１台以上の露光装置２から構成されている。なお、露光装置２はパターン転写を行う投影光学系を持つ装置である。露光装置群１は、割付処理部４および、入出力インターフェイス５および、データベース部６から構成される割付処理ステーション３に接続されている。また、露光装置群１は、半導体デバイス製造ライン全体のロット進捗情報を管理する製造管理システム７に接続されている。なお、露光処理を行う際の半導体ウエハの品種、工程、露光装置の名称および、ウエハ番号について、例えば、ウエハに形成されている製品番号を読み取ることによって入力し、製造管理システム７からネットワークを介して送信することが可能である。 また、露光装置群１内の露光装置２においてロットの露光処理が行われた場合、ネットワークを介して、割付処理ステーション内のデータベース部６に露光処理を行ったロットの品種、工程、装置の名称情報や着工日時、その際のショット数、露光量、フォーカス補正値といった製造条件が蓄積される。データベース部６は、年月日毎の品種名、投入ロット数、投入ウエハ数、まとめロット数が蓄積される生産計画データベース６１および、半導体デバイスの品種、ホトリソグラフィー工程、露光装置の名称と、該当品種・工程・装置にて着工した日時、その際のショット数、露光量、フォーカス補正値といった製造条件実績が蓄積される製造条件実績データベース６２、露光装置、機種名、照度、露光量補正係数、動作特性情報、ラフ工程の平均処理枚数など露光装置の情報が蓄積される装置マスタデータベース６３、品種毎の工程名、工程タイプ（例えば、ホトリソグラフィー工程のファイン工程、ホトリソグラフィー工程のラフ工程、エッチング工程など）が蓄積される工程マスタデータベース６４から構成される。割付処理部４は、入出力インターフェイス５を介してデータベース部６や製造管理システム７から投入計画情報、製造条件実績情報、装置マスタ情報、工程マスタ情報、仕掛りロット情報を取得し、生産計画に対応した使用露光装置の割り付け処理を行う。割付結果は入出力インターフェイス５に介して製造管理システム７に送信され、割り付け結果に応じた露光装置の使用指示を行う。また、割付結果は専用の端末に表示することもできる。製造管理システム７は、半導体デバイス製造におけるホトリソグラフィー工程だけでなく全工程における製造実績（ロット名称、ウエハ枚数、品種・工程・製造装置の名称、着工日時）の蓄積や、工程フローの管理ならび、次工程の製造指示等を行っている。割付処理部４からの仕掛りロットの問い合わせに対して、現在ラインに仕掛かっているロットのロット名称、ウエハ枚数、品種名、仕掛り工程名、最初のホトリソグラフィー工程で使用した露光装置といった情報を回答する。
【００３０】
図１７は本発明の第２の実績形態を示すブロック図である。なお、図１６と等価な部分には符号を付してある。この実施形態では、データベース部６および、割付処理部４から構成される割付処理ステーション３が製造管理システム７内に組み込まれており、各処理が実施される。
【００３１】
【発明の効果】
本発明は、半導体デバイス製造ラインにおいて、現時点の仕掛りロットとこれから投入するロットの負荷を最初のホトリソグラフィー工程で使用する露光装置毎に計算し、露光装置間の負荷ばらつきが最小となるように最初のホトリソグラフィー工程で使用する露光装置を割り付ける生産スケジューリング機能をもつため、露光装置間の負荷を平準化され、投入ロットのＴＡＴが短縮される。また、品種毎の露光装置の処理時間を推定する機能により、現状の仕掛り状況と露光装置の処理能力を考慮した適正な「条件出し作業」を行う露光装置を選定するができ、急な生産計画の変更に対応した「条件出し作業」計画を短時間で立案することができ、製造ラインの生産量を増加させることができる。
【図面の簡単な説明】
【図１】本発明の半導体デバイス製造における生産スケジューリング方法を示すフローチャートである。
【図２】本発明における露光装置の割付処理方法を示すフローを示す図である。
【図３】半導体デバイス製造フローを示す図である。
【図４】半導体デバイス製造ラインにおける仕掛りロットの概念を示す図である。
【図５】投影縮小型露光装置の構成および露光量とフォーカス補正値とレジスト寸法の関係を示す図である。
【図６】号機限定処理における露光装置の負荷を示す図である。
【図７】生産計画情報の表示例である。
【図８】着工可能である露光装置情報の表示例である。
【図９】本発明を用いてホトリソグラフィー工程で使用する露光装置を指定した日別生産計画情報の表示例である。
【図１０】本発明を用いてホトリソグラフィー工程で使用する露光装置を指定した際の露光装置号機の負荷の表示例である。
【図１１】本発明を用いてホトリソグラフィー工程で使用する露光装置を指定した際の露光装置号機の負荷の表示例である。
【図１２】本発明を用いてホトリソグラフィー工程で使用する露光装置を指定した際の露光装置号機の負荷の表示例である。
【図１３】本発明における割付装置組合せの作成方法を示すフローチャートである。
【図１４】本発明における最適割付装置組合せの算出方法を示すフローチャートである。
【図１５】本発明における日別生産計画の作成方法を示すフローチャートである。
【図１６】本発明の第１の実施形態を示すブロック図である。
【図１７】本発明の第２の実施形態を示すブロック図である。
【符号の説明】
１…露光装置群、２…露光装置、３…割付処理ステーション、４…割付処理部、５…入出力インターフェイス、６…データベース部、７…製造管理システム。[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a production scheduling method and a production scheduling system in semiconductor device manufacturing.
[0002]
[Prior art]
In semiconductor device manufacturing, as shown in FIG. 3, a predetermined device pattern is formed by a film forming process for forming an oxide film or a metal film on a semiconductor wafer, a photolithography process for transferring a circuit pattern to a semiconductor wafer, or a chemical treatment. Job shop production methods are often used to form multilayer semiconductor devices according to the process flow while repeating many manufacturing processes such as etching processes and ion implantation processes that inject impurities into semiconductor devices and change film characteristics. . Then, the processing of each process is normally performed in a unit called a lot (generally 25 sheets for a 200 mm wafer) in which a plurality of semiconductor wafers are collected.
[0003]
The semiconductor device manufacturing line has a group of manufacturing equipment corresponding to each process such as film formation, photolithography, etching, and ion implantation processes. However, device specifications differ depending on the product type and process, and there are machine differences between devices in the same manufacturing device group. Therefore, when producing a new product, the manufacturing conditions for each process and each manufacturing device are usually determined by experiments. It is necessary to perform “conditioning work” for optimization. In particular, in a process where the device specifications required for the process are strict, only the manufacturing apparatus in which this “conditioning operation” has been performed can be used for processing. For example, as shown in FIG. 5, in an exposure apparatus used in a photolithography process for transferring a circuit pattern to a resist film, which is a photosensitive material on a semiconductor wafer, a portion called a reticle that blocks light and light The circuit pattern formed by the passing portion is reduced and transferred to the resist film on the semiconductor wafer placed on the stage via the reduction lens by the exposure light irradiated from the illumination system. The exposure unit is referred to as a “shot”, and an exposure process is performed by driving the same circuit pattern on a plurality of “shots” in a semiconductor wafer. At this time, the resist dimension on the semiconductor wafer changes depending on the exposure amount as exposure light energy irradiated from the exposure apparatus and the focus correction amount between the stage and the reduction lens. Therefore, it is necessary to perform a “conditioning operation” for determining an appropriate exposure amount and focus correction value for each type, process, and manufacturing apparatus so as to obtain a predetermined resist size.
[0004]
Normally, in job shop production such as semiconductor device manufacturing, as shown in FIG. 4, a certain amount of in-process lots waiting for the next processing is held for each manufacturing device group corresponding to each process (job), Production is controlled so that the processing capacity of each manufacturing device group is maximized. On the other hand, increasing the number of in-process increases the TAT (Turn Around Time, the time from the start of production to the end) of the product, so it is necessary to generate a production schedule that takes into account the production volume and TAT.
[0005]
The following is a method for generating a production schedule for job shop production.
1) Scheduling solution (analytical method)
Using the method of operation research, formulate the schedule problem and calculate the optimal solution that maximizes the evaluation index. In general, a method is used in which input lots are allocated so that the load is leveled to a manufacturing apparatus that can input the load corresponding to the production plan. In addition, the bottleneck process of semiconductor device manufacturing is often a photolithography process that uses an exposure apparatus in which the number of apparatuses is limited because the apparatus price is high.
2) Dispatch rules (heuristic method)
A schedule is generated using a dispatch rule in which the knowledge and experience of a person engaged in the field are ruled. For example, in the dispatch rule of delivery date priority, the lot closest to the delivery date among the lots in progress for a certain process becomes the next processing target. Generally, a plurality of dispatch rules are used in combination, a score is assigned to each dispatch rule, and processing is performed from the lot with the highest score.
[0006]
[Problems to be solved by the invention]
In a semiconductor device manufacturing line that is a low-volume production of a variety of products represented by system LSI manufacturing, short TAT is a differentiating factor of other companies. In a product in which the overlay accuracy between each layer of a semiconductor device is severe, it is necessary to remove as much as possible the overlay error caused by the aberration and distortion of the reduction lens of the exposure apparatus used in the photolithography process. Therefore, “No. machine limited start” is used in which the main photolithography process (hereinafter referred to as “fine process”) is started with the same exposure apparatus. On the other hand, in a photolithography process (hereinafter, rough process) in which the required overlay accuracy is not so strict, there is no particular limitation on the exposure apparatus to be used. Generally, one product has about 30 to 40 photolithography processes, and nearly half of them are fine processes. In such “unit limited start”, the exposure apparatus used in the first photolithography process determines the apparatus to be used in the remaining fine processes. For this reason, as shown in FIG. 6, depending on the selection method of the exposure apparatus used in the first photolithography process, load variation among the exposure apparatuses occurs. For this reason, the product TAT differs depending on the exposure apparatus used in the first photolithography process, which is a bottleneck for realizing a short TAT in multi-product small-volume production. Therefore, problems in conventional production scheduling when optimizing the exposure apparatus used in the first photolithography process are described below.
1) Scheduling solution
The variety line handles hundreds of varieties. An exposure apparatus used in the photolithography process of the bottleneck process needs to perform a “conditioning operation” for optimizing manufacturing conditions in advance for each type, process, and apparatus. At this time, because there is a machine difference between exposure apparatuses, even when the exposure apparatus is different even in the same product type / process, the processing time of the lot is different. In addition, since the processing time varies depending on the type and process, it is difficult to grasp in advance the processing capability of the apparatus that is scheduled to perform the “conditioning operation”. For this reason, the scheduling result varies depending on the implementation status of the “conditioning work” for each type and process. Therefore, production scheduling including optimization of an exposure apparatus that performs “conditioning work” is required.
2) Dispatch rules
In the scheduling based on a simple dispatch rule, lots are allocated according to the vacant state of the exposure apparatus group at a certain moment. Therefore, in the long term, the load between exposure apparatuses varies as shown in FIG.
[0007]
Therefore, in order to achieve low TAT in a high-mix low-volume semiconductor device manufacturing line, the present invention instructs the appropriate “conditioning operation” of the exposure apparatus used in the photolithography process, and reduces the load between the exposure apparatuses. The present invention provides a production scheduling method and a production scheduling system in which an input lot is assigned to an exposure apparatus used in an initial photolithography process so as to be leveled.
[0008]
[Means for Solving the Problems]
The present invention includes a first database section in which production plan information such as product name, number of input lots, number of input wafers, and number of batches for each semiconductor device manufacturing is stored, product type of semiconductor device, photolithography process, The name of the exposure device, the date and time when the corresponding product / process / device was started, the number of shots at that time, the amount of exposure, the amount of exposure, the manufacturing conditions such as the focus correction value, the second database part, the exposure device, the model name, A third database section that stores information about the exposure apparatus, such as illuminance, exposure amount correction coefficient, operation characteristic information, and average number of processed rough processes, and process name and process type for each product type (for example, fine process of photolithography process) A fourth database in which process information such as rough process of photolithography process, etching process, etc.) is accumulated, and semiconductor data Using a manufacturing management system that manages the progress of lots in a chair manufacturing line, and input plan information, manufacturing condition results information, exposure apparatus information, process information, and in-process lot information acquired from the manufacturing management system, obtained from each database unit And an allocation processing unit that performs allocation processing for the exposure apparatus used in the first photolithography process corresponding to the input plan.
[0009]
The allocation processing unit is composed of the following 10 steps.
(1) obtaining a production plan to be assigned, and
(2) acquiring in-process lot information of the current production line;
(3) a step of acquiring an exposure apparatus capable of starting construction for each product type / process from manufacturing condition result information;
(4) A step of estimating the processing time of the lot for each type, process, and exposure apparatus using the manufacturing condition result information and the exposure apparatus information;
(5) calculating an initial load for each exposure apparatus using the processing time calculated in (4);
(6) a step of setting parameters necessary for the assignment instruction of the exposure apparatus;
(7) generating an allocation apparatus combination using the startable exposure apparatus information acquired in (3) and the allocation parameter information acquired in (6);
(8) calculating an optimum combination from the allocation device combinations generated in (7);
(9) outputting a production plan in which an exposure apparatus to be used in the first photolithography process is assigned in accordance with the production plan acquired in step (1) based on the assignment device combination calculated in (8); ,
(10) A step of calculating an apparatus load in the production plan created in (9) and outputting a calculation result.
[0010]
By sequentially executing the steps (1) to (10), production scheduling is performed in which an exposure apparatus used in the first photolithography process is allocated.
[0011]
DETAILED DESCRIPTION OF THE INVENTION
Next, embodiments of the present invention will be described with reference to the drawings.
[0012]
FIG. 1 is a flowchart showing an exposure apparatus allocation method in production scheduling according to the present invention. First, in step 101, using the processing time estimation information for each product type / process / exposure apparatus, the information on the lot in progress on the current semiconductor device manufacturing line, and the lot information acquired from the production plan to be allocated, A lot load is calculated for each exposure apparatus used in the first photolithography process, and an exposure apparatus used in the first photolithography process is assigned so as to minimize the load variation between the exposure apparatuses. In step 102, an exposure apparatus to be used in the first photolithography process is instructed based on the production plan that has been assigned in step 101. In step 103, instructions are given for an exposure apparatus for which “conditioning work” has not been performed in the production plan that has been assigned in step 101.
[0013]
FIG. 2 is a flowchart showing details of the allocation processing in step 101 of FIG. First, in step 201, production plan data to be subjected to exposure apparatus allocation processing is acquired. For example, production plan data can be obtained by searching from an accumulated database. Further, the operator may create production plan data directly. The production plan data includes the input date for each product type, the number of input lots, and the number of batches when a plurality of lots are input collectively. FIG. 7 shows a display example of production plan data. In this example, the target period of the allocation process is 10 days from 1/1 to 1/10, and the input varieties are three varieties A to C. The number of input lots per day, the total value of input lots for each product type in the target period, and the total number of lots when a plurality of lots are input in a line are displayed. Any data can be changed on this screen.
[0014]
Next, in step 202, information on lots in progress on the semiconductor device manufacturing line is acquired. In-process lot information can be acquired from a production management system that sequentially manages the progress of production lots. The in-process lot information includes the name of the corresponding lot, the number of wafers in the lot, the name of the product type, the name of the process currently in progress, and information on the number of the exposure apparatus used in the first photolithography process. Next, in step 203, the number information of the exposure apparatus that can start construction for each type and process is acquired. An exposure apparatus that can start construction can be acquired from manufacturing specification information that defines manufacturing conditions for each type and process. In addition, the search results for the types, processes, and exposure devices that were started within a certain period from the database that manages the history of product types, process names, start dates, etc. The device can also be calculated. FIG. 8 shows a display example of the machine information of the exposure apparatus that can start construction for each type and process. In this example, the type A has five photolithographic processes, namely layers 1 to 5, of which layers 1, 3, and 5 are fine processes and layers 2 and 4 are rough processes. In addition, there are three exposure apparatuses # 1 to # 3, and “◯” is displayed in a process that can be started. Such information can also be displayed for other varieties by selecting the varieties.
[0015]
Next, in step 204, process information corresponding to the type acquired in steps 201 and 202 is acquired, and processing time per lot is estimated for each combination of photolithography process and exposure apparatus. The process information is acquired from a database that manages process information for each product type. The process information includes process name and process type (fine process of photolithography process, rough process of photolithography process, etching process, and the like). The processing time per lot is expressed as a function of the number of shots, the set exposure amount, and the difference information of the exposure apparatus, as shown in equation (1).
[0016]
Processing time per lot = f (number of shots, exposure amount, operating characteristics of exposure apparatus) (1)
When the number of shots increases, the number of exposures per semiconductor wafer increases. Therefore, the lot processing time increases in proportion to the number of shots. The number of shots is a parameter depending on the type. When the exposure amount increases, the exposure time per shot increases, so the lot processing time increases in proportion to the exposure amount. Usually, the number of shots and the exposure amount can be obtained from a manufacturing specification that defines the manufacturing conditions for each type, process, and exposure apparatus. In addition, the number of shots and exposure amount at that time are included in the history such as the product name, process name, and start date / time at the start of each exposure apparatus. it can. The exposure amount of the type, process, and exposure device that has not yet been subjected to "conditioning work" and is not registered in the manufacturing specifications or manufacturing results database is the exposure amount correction coefficient of the exposure device relative to the standard exposure amount for each type and process. It can be calculated by multiplying. Here, the reference exposure amount for each type / process is searched for exposure amount actual data of the same or similar type, process, and exposure apparatus, and a process for removing the difference component of the corresponding exposure apparatus for each search data is performed. After that, the average value of all data can be substituted. The exposure amount correction coefficient is represented by a ratio between exposure units of an exposure amount necessary to make the resist dimension formed from the same pattern of the same reticle the same value. As a result, it is possible to estimate the lot processing time in the type, process, and exposure apparatus for which “conditioning work” is not performed. The operation characteristic data of the exposure apparatus includes illuminance irradiated from the illumination system, stage feed time for each shot, various measurement times inside the exposure apparatus, time required for setup work, and the like. For example, when there is a difference in illuminance between exposure apparatuses, the time required to irradiate the same exposure amount varies. Therefore, it is necessary to acquire illuminance data in advance for each exposure apparatus and store it in a database. In addition, the stage feed time, measurement time, and setup time are not very different between the exposure units of the same model, but the performance of the device differs depending on the model. And need to be accumulated in the database.
[0017]
In step 205, the initial load for each unit of the exposure apparatus is calculated using the lot information on the semiconductor device manufacturing line acquired in step 202. The load calculation method will be described by taking the process shown in FIG. 6 as an example. In this example, the number of photolithography processes is five. Among them, since the layers 1, 3, and 5 are fine processes, the exposure apparatus used in the layer 1 must be used in the layers 3 and 5. It is known from the in-process lot information that the in-process lot a in the layer 2 has started the layer 1 with the exposure apparatus # 2. In this case, a load for the processing time of the remaining fine process layers 3 and 5 is added to the load of the exposure apparatus # 2. At this time, the average load per day can be calculated by dividing the sum of the processing times of the fine process by the remaining TAT. In addition, the setup time may vary depending on the number of products that have been started in the exposure apparatus, and a setup correction coefficient represented by a function of the number of products introduced for each exposure apparatus may be added to the load. In addition, since the in-process lot b in the layer 3 is the fine process, the layer 1 is started by the exposure apparatus # 1. At this time, the load corresponding to the remaining fine process layers 3 and 5 is added to the load of the exposure apparatus # 1. In this way, the same processing is performed for all in-process lots, and the initial load of the fine process for each exposure apparatus is calculated. Further, the rough process load calculation includes a method of evenly distributing the load to an exposure apparatus capable of starting construction. However, in practice, in the rough process, the exposure apparatus to be used is changed according to the load condition of the exposure apparatus. Therefore, a method may be used in which the total number of rough process processes per day is calculated from the in-process lot and the production plan, and the load of the rough process is distributed to each exposure apparatus according to the average number of processed rough processes of each exposure apparatus. .
[0018]
Next, in step 206, various parameters necessary for the allocation calculation are adjusted. The adjustment work includes manual work by an operator and automatic adjustment. As an adjustment parameter, designation of a priority allocation device, designation of a use state, designation of an allocation method, and the like are performed. In the designation of the priority assignment apparatus, an exposure apparatus to be preferentially used can be designated in a specific product / process. In addition, the use state can be specified based on whether or not the exposure apparatus can be used in a specific product type and process, regardless of the information on the exposure apparatus that can be started from the manufacturing specification. In addition, it is possible to specify whether or not the entire exposure apparatus can be used, such as when a specific exposure apparatus fails. In the designation of the allocation method, the allocation process is normally performed based on exposure apparatus information that can be started from a manufacturing specification. However, when the production amount of a certain product type increases, it is necessary to increase the number of new devices to be used in addition to the past device results. In the case of such “additional designation”, the model information of the exposure apparatus to be newly used and the number of additional apparatuses are designated. In addition, although there is no equipment record for the newly introduced product, it is necessary to secure some new equipment. In the case of such “new designation”, the designation of the model information of the exposure apparatus to be used and the number of apparatuses are designated.
[0019]
Next, in step 207, an allocatable device combination for each input target product type is generated. Detailed processing will be described in detail later with reference to FIG. Then, in step 208, the optimum allocation combination is calculated, and the combination of apparatuses with the most equalized load between the exposure apparatuses is extracted. Here, a large-scale production plan such as how many lots are put into which exposure apparatus for each type is determined. Detailed processing will be described later in detail with reference to FIG. Next, in step 209, an exposure apparatus to be used for each daily production plan is assigned based on the input lot number information for each product type / exposure apparatus calculated in step 208. In the assignment method, on a certain introduction date, devices are assigned sequentially to devices with the lowest device load among allocatable devices. However, if the number of input lots exceeds the type / exposure device, the corresponding device is excluded from the input target. Detailed processing will be described in detail later with reference to FIG. FIG. 9 shows a display example of the daily production plan. In this example, the target period of the allocation process is 10 days from 1/1 to 1/10, the input types are three types A to C, and the exposure apparatuses are three units # 1 to # 3. The number of input lots for each exposure apparatus is displayed. The status of “conditioning work” is displayed for each type and exposure apparatus. “Done” is a device that has already been subjected to “Conditioning work”, “Necessary” is a device that will require “Conditioning work” in the future, and “Unfinished” is a device that has not yet undergone “Conditioning work” is there
In step 210, the load of the exposure apparatus after the input is calculated based on the daily production plan created in step 209. As with the initial load calculation in step 205, the load calculation method adds the value obtained by dividing the sum of the processing times of each fine process by the total TAT and the number of input lots to the load of the corresponding exposure apparatus. To do. At this time, the setup time may vary depending on the number of products that have been started in the exposure apparatus, and a setup correction coefficient represented by a function of the number of products introduced for each exposure apparatus may be added to the input load. FIG. 10 shows a display example of the device load after the input. In this example, there are three exposure apparatuses # 1 to # 3. For each exposure apparatus, the rough process load, the input lot fine process load, and the in-process lot fine process load are displayed. Further, when the total value of each load for each exposure apparatus exceeds the reference value, an alarm is displayed. Further, by selecting the exposure apparatus and the load category, the load for each type as shown in FIG. 11 is displayed. Further, by selecting the product type and the load category, the load for each device as shown in FIG. 12 is displayed. In step 211, the post-loading load is determined. This determination includes manual determination by an operator and automatic determination. If it is determined as “OK”, the daily production plan is fixed. If it is determined as “NG”, the process returns to step 206 to adjust the allocation parameter, and the processing from step 207 is executed. The criterion for automatic determination is that the load on each exposure apparatus is below the reference value. In the allocation parameter adjustment, the type having the largest load amount is extracted from the input loads of the exposure apparatus exceeding the reference value. When the allocation method of the product type is “normal”, the allocation method is changed to “additional designation”, the number of additional devices is set to 1, and the allocation calculation is performed again. When the allocation method is “additional designation”, the number of additional devices is increased by one and the allocation calculation is performed again. When the allocation method is “new designation”, the number of selected devices is increased by one and the allocation calculation is performed again. In this way, the processing from step 107 to step 112 is repeated until the load after the exposure apparatus is turned on becomes below the reference value.
[0020]
FIG. 13 is a flowchart showing generation of assignable device combinations in step 207 of FIG. First, at step 301, the exposure apparatus information that can be started for the first photolithography process of the corresponding product obtained at step 203 of FIG. 2, the priority assignment apparatus instruction information, the use state instruction information, and the assignment method instruction obtained at step 206 are shown. Based on the information, a selectable device list and a prioritized device list are created for each input product type. When the allocation method is “normal”, a priority allocation device list is created from the priority allocation device instruction information. Then, a selectable device list is created by excluding the devices registered in the priority assignment device list from the startable exposure device information and the use state instruction information. When the selection allocation method is “additional designation”, a priority allocation apparatus list is created from the applicable type of startable exposure apparatus and priority allocation apparatus instruction information. Then, among the machines corresponding to the model of the designated exposure apparatus, apparatuses excluding the apparatuses registered in the priority assignment apparatus list are added to the selectable apparatus list. When the selection assignment method is “new designation”, a priority assignment device list is created from the priority assignment device instruction information of the corresponding product type. Then, among the machines corresponding to the model of the designated exposure apparatus, apparatuses excluding the apparatuses registered in the priority assignment apparatus list are added to the selectable apparatus list. At this time, the same device is not included in the priority assignment device list and the selectable device list.
[0021]
Next, in step 302, an allocation device combination is generated from the priority allocation device list and the selectable device list. For example, the priority assignment device list is # 1, 3 and the selectable device list is # 2, 4, 5 as an example. First, combinations of selectable devices are generated according to the number of used devices. In this case, when using three devices: one way, when using two devices: three ways, when using one device: three ways, a total of seven ways. By adding a priority allocation device to this selectable device combination, an allocation device combination is generated. In this example, the following seven combinations are generated.
[0022]
No. 1: # 1, # 3, # 2, # 4, # 5 (priority: 2 units, selection: 3 units)
No. 2: # 1, # 3, # 2, # 4 (priority: 2 units, selection: 2 units)
No. 3: # 1, # 3, # 2, # 5 (priority: 2 units, selection: 2 units)
No. 4: # 1, # 3, # 4, # 5 (priority: 2 units, selection: 2 units)
No. 5: # 1, # 3, # 2 (priority: 2 units, selection: 1 unit)
No. 6: # 1, # 3, # 4 (priority: 2 units, selection: 1 unit)
No. 7: # 1, # 3, # 5 (priority: 2 units, selection: 1 unit)
FIG. 14 is a flowchart showing a method for calculating the optimum allocation combination in step 208 of FIG. First, in step 401, a process is defined in which steps 402 to 405 are repeated by the number of allocation device combinations calculated in step 207 of FIG. Next, in step 402, the input lot is distributed for each allocation device list in a certain allocation possible combination No. Combination No. above A method for allocating the input lot of product type A will be described by taking 1 as an example. In this example, a certain type A is assigned to five exposure apparatuses # 1 to # 5 (m_total). At this time, the initial load of each exposure apparatus is A (m). Here, m is an integer corresponding to the exposure apparatuses # 1 to # 5. Further, the total number of input lots is 20 (X_total), and the total number of lots input to each exposure apparatus is X (m). Actually, since lots are collectively put into the line, it is necessary to perform distribution processing with the number of batches considering the number of batches, but here the number of batches is set to 1. At this time, the equation (2) holds for X (m).
[0023]
X (1) + X (2) + X (3) + X (4) + X (5) = 20 (2)
When this is generalized, equation (3) is obtained.
[0024]
[Expression 1]

[0025]
Further, since the processing time varies depending on the exposure apparatus even for the same type, the load on the input lot differs. Therefore, a value obtained by dividing the sum of the processing times of each fine process of each exposure apparatus by the total TAT is set as a load coefficient a (m) for the input lot. The load B (m) of each exposure apparatus after being charged is expressed by equation (4).
[0026]
B (m) = A (m) + a (m) ・ X (m) (4)
It is only necessary to determine X (m) as the number of input lots of each exposure apparatus so that the load B (m) of the exposure apparatus after input is leveled. However, at this time, the constraint condition of the expression (3) must be satisfied. As a method for solving such an optimization problem, there is a dynamic calculation method of an operation research method. In this method, X (m) that maximizes the evaluation function g (B (m)) can be obtained under the constraint condition of Equation (3).
[0027]
Next, in step 403, an evaluation index in the input load distribution result calculated in step 401 is calculated. As an evaluation index, the standard deviation or average of the load of each exposure apparatus may be used. In Step 404, when the combination No is smaller than the total number of combinations, the combination No is updated, and the process returns to Step 402. If the combination No is equal to the total number of combinations, the process proceeds to the next step. In step 405, among the evaluation indexes for each combination No calculated in step 403, the maximum combination number (or the minimum, depending on the evaluation index definition method) is extracted, and the assigned device list and each device at that time are extracted. The total number of input lots is read and the optimal allocation device combination is obtained.
[0028]
FIG. 15 is a flowchart showing a method for calculating the daily production plan in step 209 of FIG. First, in step 501, steps 502 to 507 are defined to repeat the daily device allocation process for the number of target days of the allocation process. In step 502, a process of repeating steps 503 to 506 for repeating the daily device allocation process for the number of input products to be allocated is defined. In step 503, if there is an input lot of a certain input date and type, the total number of input lots calculated in step 208 in FIG. 2 is the exposure apparatus to which the corresponding input lot can be assigned and the total number of input lots of the corresponding type / exposure apparatus. Of the smaller exposure apparatuses, the exposure apparatus that minimizes the apparatus load at that time is calculated. Next, in step 504, the exposure apparatus with the smallest apparatus load calculated in step 503 is assigned as the used exposure apparatus for the corresponding lot. In step 505, the load of the input lot assigned in step 504 is assigned to the corresponding exposure apparatus. As with the calculation of the initial load in step 205 in FIG. 2, the load calculation method is obtained by multiplying the sum of the processing times of each fine process by the total TAT and multiplying the number of input lots by the value of the corresponding exposure apparatus. Add to load. At this time, the setup time may differ depending on the number of products that have been started in the exposure apparatus, and the setup correction coefficient represented by a function of the number of products introduced for each exposure apparatus may be added to the input load. Further, in order to actually put lots into the line, the exposure apparatus is assigned in batch units considering the number of batches. For example, if the number of input lots of a product A on a certain day = 12 and the number of batches = 4, the number of batches is 3, so the calculation from step 503 to step 505 is performed in three steps. In step 506, if the product number is smaller than the number of input products, the product number is updated, and the process returns to step 503. If the product number is equal to the number of input products, the process proceeds to the next step. In step 507, if the input date is smaller than the end date of the allocation process, the input date is set as the update resource, and the process returns to step 502. If the input date is equal to the end date of the allocation process, the process ends.
[0029]
FIG. 16 is a block diagram showing the overall configuration of the first embodiment of the present invention. In the figure, the exposure apparatus group 1 is composed of at least one exposure apparatus 2. The exposure apparatus 2 is an apparatus having a projection optical system that performs pattern transfer. The exposure apparatus group 1 is connected to an allocation processing station 3 including an allocation processing unit 4, an input / output interface 5, and a database unit 6. The exposure apparatus group 1 is connected to a manufacturing management system 7 that manages lot progress information of the entire semiconductor device manufacturing line. The semiconductor wafer type, process, exposure apparatus name, and wafer number when performing the exposure process are input by, for example, reading the product number formed on the wafer, from the manufacturing management system 7 via the network. Can be transmitted. In addition, when lot exposure processing is performed in the exposure apparatus 2 in the exposure apparatus group 1, the type, process, and name of the lot for which exposure processing has been performed on the database unit 6 in the allocation processing station via the network. Manufacturing conditions such as information, start date / time, number of shots at that time, exposure amount, and focus correction value are accumulated. The database unit 6 includes a production plan database 61 in which the name of each product type, the number of input lots, the number of input wafers, and the total number of lots are accumulated, the name of the semiconductor device type, the photolithography process, the exposure apparatus, and the corresponding type.・ Manufacturing condition results database 62 that accumulates manufacturing condition results such as the date and time of start of the process / apparatus, the number of shots at that time, exposure amount, and focus correction value, exposure apparatus, model name, illuminance, exposure amount correction coefficient, operation An apparatus master database 63 in which information of exposure apparatuses such as characteristic information and the average number of processed rough processes is stored, process name and process type for each product type (for example, fine process of photolithography process, rough process of photolithography process, etching process) Etc.) is stored in the process master database 64. The allocation processing unit 4 acquires the input plan information, manufacturing condition result information, device master information, process master information, and in-process lot information from the database unit 6 and the manufacturing management system 7 via the input / output interface 5, and creates the production plan. The corresponding used exposure apparatus is assigned. The allocation result is transmitted to the manufacturing management system 7 via the input / output interface 5 and the use of the exposure apparatus is instructed according to the allocation result. In addition, the allocation result can be displayed on a dedicated terminal. The manufacturing management system 7 accumulates manufacturing results (lot name, number of wafers, name of product / process / manufacturing equipment, start date / time) in all processes as well as photolithography processes in semiconductor device manufacturing, process flow management, It gives instructions for manufacturing the next process. In response to an in-process lot inquiry from the allocation processing unit 4, information such as the lot name of the lot currently in process on the line, the number of wafers, the product name, the in-process process name, and the exposure apparatus used in the first photolithography process Answer.
[0030]
FIG. 17 is a block diagram showing a second performance mode of the present invention. In addition, the code | symbol is attached | subjected to the part equivalent to FIG. In this embodiment, an allocation processing station 3 including a database unit 6 and an allocation processing unit 4 is incorporated in the manufacturing management system 7 and each process is performed.
[0031]
【The invention's effect】
In the semiconductor device production line, the present invention calculates the load of the current in-process lot and the lot to be loaded for each exposure apparatus used in the first photolithography process so that the load variation between the exposure apparatuses is minimized. Since it has a production scheduling function for assigning an exposure apparatus to be used in the first photolithography process, the load between the exposure apparatuses is leveled and the TAT of the input lot is shortened. In addition, with the function to estimate the processing time of the exposure equipment for each product type, it is possible to select an exposure equipment that performs appropriate “conditioning work” considering the current work in progress and the processing capacity of the exposure equipment. The “conditioning work” plan corresponding to the plan change can be made in a short time, and the production amount of the production line can be increased.
[Brief description of the drawings]
FIG. 1 is a flowchart showing a production scheduling method in manufacturing a semiconductor device according to the present invention.
FIG. 2 is a flowchart showing an allocation processing method for an exposure apparatus according to the present invention.
FIG. 3 is a diagram showing a semiconductor device manufacturing flow.
FIG. 4 is a diagram showing a concept of a work lot in a semiconductor device manufacturing line.
FIG. 5 is a diagram showing a configuration of a projection reduction type exposure apparatus and a relationship among an exposure amount, a focus correction value, and a resist dimension.
FIG. 6 is a view showing a load of the exposure apparatus in the number-limited process.
FIG. 7 is a display example of production plan information.
FIG. 8 is a display example of exposure apparatus information that can be started.
FIG. 9 is a display example of daily production plan information specifying an exposure apparatus used in a photolithography process using the present invention.
FIG. 10 is a display example of a load of an exposure apparatus number when an exposure apparatus used in a photolithography process is specified using the present invention.
FIG. 11 is a display example of a load of an exposure apparatus number when an exposure apparatus used in a photolithography process is specified using the present invention.
FIG. 12 is a display example of a load of an exposure apparatus number when an exposure apparatus used in a photolithography process is specified using the present invention.
FIG. 13 is a flowchart showing a method of creating an allocation device combination according to the present invention.
FIG. 14 is a flowchart showing a method for calculating an optimum allocation device combination in the present invention.
FIG. 15 is a flowchart showing a method for creating a daily production plan according to the present invention.
FIG. 16 is a block diagram showing a first embodiment of the present invention.
FIG. 17 is a block diagram showing a second embodiment of the present invention.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Exposure apparatus group, 2 ... Exposure apparatus, 3 ... Assignment processing station, 4 ... Assignment processing part, 5 ... Input / output interface, 6 ... Database part, 7 ... Manufacturing management system.

Claims (4)

In producing a semiconductor device using a plurality of exposure apparatuses, a semiconductor device production scheduling method for assigning an exposure apparatus used in an initial exposure process,
The allocation process uses the operation characteristic information including the number of shots for each type of the semiconductor device and the exposure amount set for each type / process / exposure apparatus and illuminance information of the exposure apparatus, for each type of the semiconductor device. The process of estimating the processing time per lot of each exposure process or for each exposure apparatus, and the load of multiple exposure processes subject to the exposure apparatus used in the initial exposure process of the in-process lot Restrictions on the exposure apparatus used in the subsequent first exposure process from the combination of the processing calculated for each exposure apparatus used in the exposure process and the exposure apparatus used in the first exposure process corresponding to the production plan information of the semiconductor device. This includes a process for assigning the exposure apparatus used in the first exposure process so that the variation in load in the plurality of exposure processes received is minimized among the exposure apparatuses. Production scheduling method of a semiconductor device according to claim.   In the processing time estimation processing per lot in claim 1, when there is no exposure amount set for each type / process / exposure apparatus, the exposure amount actual data of the same or similar type, process, and exposure apparatus is searched. Then, exposure is performed by multiplying the exposure amount correction coefficient of the exposure apparatus corresponding to the reference exposure amount calculated by the average value of all data after performing the process of removing the machine difference component of the exposure apparatus corresponding to each search data. A production scheduling method for semiconductor devices, characterized in that quantity estimation processing is performed. In the exposure apparatus allocation process according to claim 1 , not only an exposure apparatus having a track record of starting construction of a combination of allocation target apparatuses in the past, but also including and newly generating an exposure apparatus to be used A production scheduling method for a semiconductor device, characterized in that presence / absence of a start of construction is displayed in an allocation result of an exposure apparatus. A first database section for storing production plan information in semiconductor device manufacturing; a second database section for storing actual manufacturing condition results for the exposure process; and a third database section for storing exposure apparatus characteristic information. A fourth database unit for storing process information, a manufacturing management system for managing a lot progress status in the semiconductor device manufacturing line, input plan information acquired from each database unit, manufacturing condition result information, exposure apparatus information, and process information and using the work-in-progress lot information acquired from the manufacturing control system, consists allocation processing section which performs allocation only processing of an exposure apparatus used in the first exposure process corresponding to the input plan,
The allocation processing unit uses the operation characteristic information including the number of shots for each type of the semiconductor device and the exposure amount set for each type / process / exposure apparatus and the illuminance information of the exposure apparatus, for each type of semiconductor device. Estimate the processing time per lot, or for each exposure process, or for each exposure apparatus, and load the load of multiple exposure processes subject to the exposure apparatus used in the initial exposure process of the in-process lot. A plurality of exposures that are calculated for each exposure apparatus used in the above and subject to the restrictions of the exposure apparatus used in the subsequent first exposure process from the combination of the exposure apparatuses used in the first exposure process corresponding to the production plan information of the semiconductor device Production of semiconductor devices characterized by assigning the exposure apparatus used in the first exposure process so that the variation in load in the process is minimized among the exposure apparatuses Scheduling system.

Images