SEMICONDUCTOR MANUFACTURING APPARATUS, SEMICONDUCTOR MANUFACTURING FACILITY AND SEMICONDUCTOR MANUFACTURING METHOD

Abstract

For a semiconductor manufacturing apparatus provided with a plurality of processors and a number of loader-unloaders exceeding them, the apparatus includes a plurality of processors (11A and 11B) and a number of loader-unloaders (15X, 15Y and 15Z) exceeding them in order to reduce the manufacturing time, even if the processor for processing is designated for each manufacturing batch, and the processor for processing transports an object to be processed (WF) designated for each manufacturing batch into the loader-unloader in a unit of the manufacturing batch, each object to be processed is transported to the processor in a single unit and processed and then transported to the loader-unloader.In this semiconductor manufacturing apparatus (1), the production batch of the object to be processed is transported by each processor into each loader-unloader while leaving the remaining loader-unloader empty, after which each of the processors starts processing, and while the object to be processed by each processor, the processor that finishes the processing relatively early is predicted for each processor based on production information of the object to be processed, and before finishing the processing by each processor, the production batch to be processed by the processor that was predicted to finish the processing relatively early is transported into the loader-unloader that is currently empty.

Description

AREA OF INVENTION

[0001] The present invention relates to a semiconductor manufacturing device, a semiconductor manufacturing plant and a semiconductor manufacturing process. BACKGROUND OF THE INVENTION

[0002] For a semiconductor manufacturing device equipped with multiple processing chambers and capable of setting multiple transport control patterns to the processing chamber, a control system of the semiconductor manufacturing device is known which collects data relating to the current operating state and the current stock batch, data relating to the wafer processing time tailored to a recipe specifying the processing content for each batch to be processed by the semiconductor manufacturing device, and collects any time loss that occurred during the processing of each batch unit, calculates the average holding time of all stock batches for each wafer transport control content based on each of the collected data, and selects the wafer transport control content based on the magnitude relationship of each calculated average holding time (see patent reference 1).Accordingly, even if there is a time loss during batch treatment, there is a way to instruct the semiconductor manufacturing equipment to perform the wafer transport control treatment appropriately. STATE OF THE ART Patent literature

[0003] Patent literature 1: Japanese patent disclosure publication no. 2010-251507 SUMMARY OF THE INVENTION: Problems that the invention is intended to solve

[0004] However, if the processing chamber for each production batch in the semiconductor manufacturing facility is designated according to customer specifications and the like, the conventional technology described above cannot be used.

[0005] The present invention aims to solve the problem of providing a semiconductor manufacturing apparatus, a semiconductor manufacturing plant and a semiconductor manufacturing process that can reduce the manufacturing time even when the number of processors to be processed for each production batch is determined, wherein the semiconductor manufacturing apparatus is provided with a plurality of processors and a number of loaders / dischargers that exceeds this. Means for solving the tasks

[0006] The present invention provides a semiconductor manufacturing apparatus equipped with a plurality of processors and a number of loaders / unloaders exceeding this plurality, wherein the processor transports an object to be treated, which is determined for each production batch, into the loader / unloader in a unit of the production batch, wherein each object to be treated is transported to the processor in a single unit and treated, and then transported to the loader / unloader.The above problems are solved by the semiconductor manufacturing device, which transports the production batch of the object to be treated through each processor into each loader-unloader, while the remaining loader-unloader is left empty, whereupon each of the processors begins the treatment; and while the object to be treated is being treated by each processor, the processor that will finish the treatment relatively early is predicted for each processor based on the production information of the object to be treated; and before each processor finishes the treatment, the production batch to be treated by the processor that was predicted to finish the treatment relatively early is transported into the loader-unloader that is currently empty.

[0007] In addition, the present invention provides a semiconductor manufacturing process which uses a semiconductor manufacturing apparatus equipped with a plurality of processors and a number of loaders / unloaders exceeding the plurality of processors, which transports by the processor to treat an object to be treated, which is determined for each production batch, into the loader / unloader in a unit of the production batch, then transports to the processor and treats each object to be treated in a single unit and then transports each object to be treated to the loader / unloader.The above problems are solved by the semiconductor manufacturing process, which begins treatment by each processor after the manufacturing batch of the object to be treated has been transported by each processor into each loader-unloader, while the remaining loader-unloader is left empty; predicts the processor that will end treatment relatively early while the object to be treated is being processed by each processor, based on the manufacturing information of the object to be treated for each processor; and, before each processor ends treatment, transports the manufacturing batch to be treated by the processor that was predicted to end treatment relatively early into the loader-unloader that is currently empty.

[0008] In the present invention, it is preferred that when the processor starts the treatment for a last or penultimate object to be treated of the production batch present in the loader-unloader, the processor predicts that the treatment will end relatively early.

[0009] Furthermore, in the present invention, the manufacturing information of the object to be treated can include a processing time for the object to be treated and the number of objects to be treated that are to be contained in a processing sequence or the manufacturing batch.

[0010] Furthermore, in the present invention, it is preferred that when the treatment of one production batch in one processor is completed, the treatment of another production batch by that processor is started if another production batch to be treated by that processor is present in one of the loader-unloaders.

[0011] The present invention also solves the above problems by means of a semiconductor manufacturing plant in which several semiconductor manufacturing devices according to the present invention are installed.

[0012] The present invention preferably comprises an integrated controller that aggregates the processor manufacturing information sent by the multiple semiconductor manufacturing devices and predicts the processor that will complete the processing relatively early; a production management device that, based on the prediction of the integrated controller, instructs the multiple semiconductor manufacturing devices to transport the production batch to be processed by the processor that was predicted to complete the processing relatively early into the loader-unloader, which is currently empty, before the processing by each processor is completed; and a transport device that transports the production batch to the multiple semiconductor manufacturing devices based on the instruction from the production management device. Effect of the invention

[0013] According to the present invention, before each processor in the semiconductor manufacturing apparatus, which is equipped with a plurality of processors and a greater number of loaders / unloaders, completes its processing, the production batch to be processed by the processor predicted to complete its processing relatively early is transferred to the currently empty loader / unloader. This allows the processor to manufacture continuously without an interval and reduces the time waiting for the loader / unloader to begin manufacturing. Consequently, even if the processor is designated to process each production batch, the manufacturing time for the object being processed can be reduced.In particular, when more than one semiconductor manufacturing device equipped with a variety of processors is installed in the semiconductor manufacturing plant, a large variety of semiconductor products with different specifications can be efficiently manufactured by using a large number of processors equivalent to a variety of devices, for example, leading to increased efficiency. BRIEF DESCRIPTION OF THE DRAWINGS Fig. Figure 1 is a planar block diagram illustrating a vapor phase growth device according to an embodiment of a semiconductor manufacturing device of the present invention. Fig. Figure 2 is a schematic block diagram of the vapor phase growth device of Fig. 1, which illustrates the handling of a production batch and an object to be treated. Fig. Figure 3 is a time diagram showing an example and a comparative example of the handling of the production batch and the object to be treated using the vapor phase growth device in Fig. 2 shows. Fig. 4A is a timing diagram (part 1) that shows a procedure from the example in Fig. 3 illustrated. Fig. 4B is a timing diagram (part 2) that shows the procedure of the example in Fig. 3 illustrated. Fig. 4C is a timing diagram (part 3) that shows the procedure of the example in Fig. 3 illustrated. Fig. 4D is a time diagram (part 4) that shows the procedure of the example in Fig. 3 illustrated. Fig. Figure 5 is a flowchart that shows the steps involved in handling the production batch and the object to be treated by the vapor phase growth device in Fig. 2 illustrated. Fig. Figure 6 is a block diagram illustrating an embodiment of a semiconductor manufacturing plant in which several semiconductor manufacturing devices are arranged. Fig. 1 are installed. MODE FOR EXECUTING THE INVENTION

[0014] An embodiment of the present invention is described below based on the drawings. The semiconductor manufacturing apparatus and semiconductor manufacturing method of the present invention imply an apparatus and a method for manufacturing a semiconductor apparatus and are not limited, but include an apparatus and a method for chemical vapor deposition (CVD) that deposits a thin film on a surface of a semiconductor, an etching apparatus and an etching method that create a slight unevenness on the semiconductor, a cleaning apparatus and a cleaning method that clean the surface of the semiconductor, and the like.In the embodiment described below, a vapor phase growth apparatus and a method that forms an epitaxial silicon film on a surface of a silicon single crystal wafer are cited as an example of the semiconductor manufacturing apparatus and method, and the embodiment of the present invention is described.

[0015] Fig. Figure 1 is a planar block diagram illustrating a vapor phase growth device 1 according to an embodiment of the present invention. A main body of the vapor phase growth device 1, shown in the center of the diagram, is illustrated in a top view. The vapor phase growth device 1 of the present embodiment is a so-called CVD device and is provided with: a pair of reaction furnaces 11A and 11B; a wafer transfer chamber 12, which accommodates a first robot 121 that handles a wafer WF, such as a silicon single-crystal wafer; a pair of loading lock chambers 13A and 13B; a factory interface 14, which accommodates a second robot 141 that handles the wafer WF; and loading ports 15X, 15Y, and 15Z, which accommodate a wafer storage container 15 (also referred to as a cassette housing or FOUP) that stores a plurality of wafers WF.

[0016] The factory interface 14 is a zone configured to maintain the same air atmosphere as the cleanroom of loading ports 15X, 15Y, and 15Z, into and out of the wafer storage container 15. Factory interface 14 is equipped with a second robot 141, which retrieves a pre-processing wafer WF from the wafer storage container 15 and places it in loading chambers 13A and 13B. It also retrieves a post-processing wafer WF, which is transported to loading chambers 13A and 13B within the wafer storage container 15. The second robot 141 is controlled by a second robot controller 142, and a second blade 143, mounted at the distal end of a robot hand, moves along a pre-programmed path.

[0017] First doors 131A and 131B, which can be opened and closed with an airtight seal, are provided between each of the loading chambers 13A and 13B with the factory interface 14, while second doors 132A and 132B, which can also be opened and closed with an airtight seal, are provided between each of the loading chambers 13A and 13B with the wafer transfer chamber 12. Furthermore, each of the loading chambers 13A and 13B serves as a space in which atmospheric gas exchange takes place between the wafer transfer chamber 12, which is configured to have an inert gas atmosphere, and the factory interface 14, which is configured to have an air atmosphere. Therefore, a suction device that vacuum-evacuates the interior of the loading lock chambers 13A and 13B, and a supply device that supplies inert gas to the loading lock chambers 13A and 13B, are provided.

[0018] For example, if the wafer WF is transported from the wafer storage container 15 via the loading chamber 13A to the wafer transfer chamber 12 before treatment, in a state in which the first door 131A on the side of the factory interface 14 is closed, the second door 132A on the side of the wafer transfer chamber 12 is closed and the loading chamber 13A has an inert gas atmosphere, the wafer WF is removed from the wafer storage container 15 using the second robot 141, the first door 131A on the side of the factory interface 14 is opened and the wafer WF is transported to the loading chamber 13A.Next, after the first door 131A on the side of the factory interface 14 is closed and the loading lock chamber 13A has been returned to an inert gas atmosphere, the second door 132A on the side of the wafer transfer chamber 12 is opened and the wafer WF is transported to the wafer transfer chamber 12 using the first robot 121.

[0019] Conversely, when the wafer WF is transported from the wafer transfer chamber 12 via the loading lock chamber 13A to the wafer storage container 15 after treatment, in a state in which the first door 131A on the side of the factory interface 14 is closed, the second door 132A on the side of the wafer transfer chamber 12 is closed and the loading lock chamber 13A has an inert gas atmosphere, the second door 132A on the side of the wafer transfer chamber 12 is opened and the wafer WF in the wafer transfer chamber 12 is transported to the loading lock chamber 13A using the first robot 121. Next, after the second door 132A on the side of the wafer transfer chamber 12 is closed and the loading lock chamber 13A has been returned to an inert gas atmosphere, the first door 131A on the side of the factory interface 14 is opened and the wafer WF is transported to the wafer storage container 15 using the second robot 141.

[0020] Wafer transfer chamber 12 is configured as a sealed chamber connected on one side to loading chambers 13A and 13B via the second doors 132A and 132B, which can be opened and closed and have an airtight seal, and on the other side via shut-off valves 114A and 114B, which can be opened and closed and also have an airtight seal. The first robot 121, which transports the wafer WF from loading chambers 13A and 13B to reaction chambers 111A and 111B before treatment and transports the wafer WF after treatment from reaction chambers 111A and 111B back to loading chambers 13A and 13B, is installed on wafer transfer chamber 12. The first robot 121 is controlled by a first robot controller 122, and a first blade 123, mounted at a distal end of a robot hand, moves along an operating path that has been taught in advance.

[0021] An integrated controller 16, which integrates the entire control of the vapor phase growth device 1, the first robot controller 122 and the second robot controller 142 send and receive mutual control signals.

[0022] If, in addition, an operating command signal is sent from the integrated controller 16 to the first robot controller 122, the first robot controller 122 controls the operation of the first robot 121, and an operating result from the first robot 121 is sent from the first robot controller 122 to the integrated controller 16. Accordingly, the integrated controller 16 detects an operating status of the first robot 121. Similarly, if an operating command signal is sent from the integrated controller 16 to the second robot controller 142, the second robot controller 142 controls the operation of the second robot 141, and an operating result from the second robot 141 is sent from the second robot controller 142 to the integrated controller 16. Accordingly, the integrated controller 16 detects an operating status of the second robot 141.

[0023] Wafer transfer chamber 12 is supplied with inert gas from an inert gas supply device, which is not shown in the drawings, and the gas in wafer transfer chamber 12 is cleaned by a cleaner (cleaning dust collector) connected to a suction port, after which the gas is released outside the system. Although a detailed illustration is omitted, this type of cleaner could, for example, use a conventionally known pressurized water cleaner.

[0024] The reaction furnaces 11A and 11B are devices for growing an epitaxial film on the surface of the wafer WF using the CVD process and include reaction chambers 111A and 111B; susceptors 112A and 112B, on which the wafer WF is placed and rotated, are provided within the reaction chambers 111A and 111B. Additionally, gas supply devices 113A and 113B are provided, which supply hydrogen gas and raw material gas for growing a CVD film (if the CVD film is an epitaxial silicon film, the raw material gas can be, for example, silicon tetrachloride SiCl4 or trichlorosilane SiHCl3) to the reaction chambers 111A and 111B. Although omitted from the drawings, a heat lamp is provided at the perimeter of reaction chambers 111A and 111B to raise the temperature of the wafer WF to a predetermined temperature.Furthermore, shut-off valves 114A and 114B are provided between the reaction chambers 111A and 111B and the wafer transfer chamber 12, and the airtight seal between the reaction chambers 111A and 111B and the wafer transfer chamber 12 is ensured by closing the shut-off valves 114A and 114B. Various controls, such as the actuation of the susceptors 112A and 112B of these reaction ovens 11A and 11B, the supply and stopping of gas by the gas supply devices 113A and 113B, the switching on and off of the heat lamp, and the opening and closing of the shut-off valves 114A and 114B, are controlled by a command signal from the integrated control unit 16. The in . Fig. The vapor phase growth apparatus shown in Figure 1 is an example provided with a pair of reaction furnaces 11A and 11B, but may include three or more reaction furnaces.

[0025] A cleaner (cleaning dust collector) with a similar configuration to that of wafer transfer chamber 12 is also provided for reaction furnaces 11A and 11B. In other words, hydrogen gas or raw material gas supplied by gas feed devices 113A and 113B is cleaned by the cleaner, which is connected to an exhaust port provided for reaction chambers 111A and 111B, and then released outside the system. For example, a conventional pressurized water cleaner can also be used for this purpose.

[0026] When the epitaxial silicon film is formed on the silicon single-crystal wafer using the vapor phase growth device 1, which is provided with two reaction furnaces 11A and 11B, as shown in Fig. As shown in Figure 1, depending on the product specifications of wafer WF, the reaction furnace used, either 11A or 11B, can be determined at the time of production planning. For example, if the susceptors 112A and 112B provided for reaction furnaces 11A and 11B differ according to the product specifications, reaction furnace 11A or 11B is selected for producing wafer WF based on the susceptor used. Furthermore, producing wafer WF from the same production batch in the same reaction furnace, 11A or 11B, maintains consistent quality between production batches and is also more suitable for conducting defect analysis.

[0027] Therefore, all wafers WF stored in a single wafer storage container 15 are individually transported to one of the predetermined reaction furnaces 11A and 11B for processing and then returned to the original wafer storage container 15. In other words, when the wafer storage container 15, which holds the wafers WF to be processed in reaction furnace 11A, is loaded into the loading port 15X in Fig. If the wafers are transported to reaction furnace 11A, processed, and then returned to the wafer storage container 15 at loading port 15X, all wafers WF are transported to reaction furnace 11A and processed. On the other hand, if the wafer storage container 15, which holds the wafers WF to be processed in reaction furnace 11B, is transported to loading port 15Y, then... Fig. 1 is transported, all wafers WF are transported to the reaction furnace 11B and treated and then returned to the wafer storage container 15 of the loading port 15Y.

[0028] In this way, the wafer transport control, in which the wafer WF of the production batch being transported into loading port 15X is processed only in reaction furnace 11A, and the wafer WF of another production batch being transported into a different loading port 15Y is processed only in a different reaction furnace 11B, is also referred to as the parallel operating mode. In contrast, the wafer transport control, in which the wafer WF of the production batch being transported into loading port 15X is processed sequentially in one of empty reaction furnaces 11A and 11B, and after all wafers WF of the production batch being transported into loading port 15X have been processed, the wafer of the production batch being transported into a different loading port 15Y is processed sequentially in one of empty reaction furnaces 11A or 11B, is also referred to as the serial operating mode.

[0029] In this example, if transport control is used in parallel operating mode, as in Fig. As shown in Figure 1, when the vapor phase growth device 1 comprises the number of loading ports 15A, 15B, and 15C exceeding the number of reaction furnaces 11A and 11B, that is, when the number of reaction furnaces is N, the vapor phase growth device 1 comprises (N+1) or more loading ports, the inventors of the present invention have thoroughly investigated whether the wafer storage container 15 of the production batch to be processed in one of N reaction furnaces should be transported in (N+1) or more loading ports. In this example, as shown in Figure 1, the number of loading ports 15A, 15B, and 15C exceeds the number of reaction furnaces 11A and 11B, i.e., when the number of reaction furnaces is N, the vapor phase growth device 1 comprises (N+1) or more loading ports. Fig. Figure 1 shows a case with two reaction furnaces 11A and 11B and three loading ports 15A, 15B and 15C.

[0030] Fig. Figure 2 is a schematic block diagram of the vapor phase growth device 1 of Fig. Figure 1 illustrates the handling of the wafer WF (object to be treated) from the same production batch. One in Fig. The production management device 2 shown is a device that integrally manages all manufacturing devices of a production process, including several vapor phase growth devices 1. The production management device 2 creates a production plan for a product and also manages the production plan by outputting the created production plan of the product to each manufacturing device, which includes several vapor phase growth devices 1 and a transport device 3, and by inputting progress information from each manufacturing device.The transport device 3 is a system that transports the wafer WF, which has completed the treatment in the previous step, in a state in which it is stored in the wafer storage container 15 to the next step, and based on a command from the production management device 2, the wafer storage container 15 of the predetermined production batch is transported in the previous step into one of the loading ports 15X, 15Y and 15Z of the vapor phase growth device 1 of the present embodiment.

[0031] Fig. Figure 3 is a time diagram showing an example and a comparative example of handling wafer WF of the same production batch using vapor phase growth device 1. Fig. 2 shows. The upper drawing of Fig. Figure 3 shows an example of the production plan created by the production management device 2, where "pi" indicates the sequence of the production plan, "A or B" designates reaction furnaces 11A or 11B for processing, and "quantity" indicates the number of wafers WF contained in a single production batch. For example, a first production batch comprises seven wafers WF to be processed in reaction furnace 11A, and a second production batch comprises five wafers WF to be processed in reaction furnace 11B. In this example, the production plan is that for all ten production batches, from the first to the tenth, 34 wafers WF are processed in reaction furnace 11A and 37 wafers WF are processed in reaction furnace 11B.Furthermore, the production plan, which is created by the production management device 2, is determined according to the product specifications to be manufactured as desired, and there are no special rules.

[0032] First, the inventors of the present invention investigated how productivity would turn out if the wafer storage container 15 were loaded into three loading ports 15X, 15Y and 15Z sequentially according to the sequence (pi) of the production plan shown in the upper drawing of Fig. As indicated in point 3, it is transported. As shown in the lower drawing of Fig. As shown in Figure 3, the first production batch is transported to loading port 15X, the second production batch is transported to loading port 15Y, and the third production batch is transported to loading port 15Z. The wafer WF of the first production batch, which is transported to loading port 15X, should be processed in reaction furnace 11A, and thus the processing in reaction furnace 11A is started. Simultaneously, the wafer WF of the production batch, which is transported to loading port 15Y, should be processed in reaction furnace 11B, and thus the processing in reaction furnace 11B is started. However, both reaction furnaces 11A and 11B are currently performing processing, and therefore the wafer WF of the production batch, which is transported to the remaining loading port 15Z, is in a standby state. This standby state is indicated by the dotted frame in the lower drawing. Fig. 3 shown.

[0033] In the lower drawing of Fig. At time T11, the treatment in reaction furnace 11B is completed, and the wafer storage bin 15, in which the wafer WF is stored after treatment, is transported out of loading port 15Y. This leaves reaction furnace 11B empty. However, at this time, the production batch in loading port 15Z, which is waiting to be treated, is a production batch that should be treated in reaction furnace 11A, not reaction furnace 11B. Therefore, in its empty state, reaction furnace 11B has no option but to transport the wafer storage bin 15, in which the fourth production batch is stored, to loading port 15Y. And after the wafer storage bin 15, in which the fourth production batch is stored, is transported to loading port 15Y, the wafer WF of this production batch is transported to reaction furnace 11B and treated.At time T11, the wafer WF of the production batch in loading port 15Z must remain in the ready state, resulting in a time loss. Furthermore, the reaction furnace 11B cannot start the treatment until the next wafer storage bin 15 is transported into loading port 15Y in the empty state, and this also leads to a time loss. This time loss also occurs in the fifth production batch at time T12 and in the seventh production batch at time T13, as shown in the lower drawing. Fig. 3 shown.

[0034] In other words, if the wafer storage container 15 is transported sequentially to the three loading ports 15X, 15Y and 15Z following the sequence (pi) of the production plan, depending on the end time of the treatment in the two reaction furnaces 11A and 11B, time losses can occur at times T11 (third production batch), T12 (fifth production batch) and T13 (seventh production batch), causing a decrease in productivity.

[0035] In contrast, the vapor deposition device 1 of the present embodiment transports the production batch of wafer WF, which is to be treated in each of the reaction furnaces 11A and 11B, into each of the loading ports 15X and 15Y, while the remaining loading port 15Z is left empty without being transported into the wafer storage container 15, whereupon each of the reaction furnaces 11A and 11B begins the treatment. Next, during the treatment of the wafer WF in each of the reaction furnaces 11A and 11B, a prediction is made, based on the production information of the wafer WF in each of the reaction furnaces 11A and 11B, as to which reaction furnace 11A or 11B will terminate the treatment relatively early.Then, before the treatment is completed in each of the reaction furnaces 11A and 11B, the wafer storage container 15 of the production batch to be treated in the reaction furnace 11A or 11B, which is predicted to complete the treatment relatively early, is transported to the loading port 15Z, which is currently empty.

[0036] In this example, the timing for predicting which reaction furnace 11A or 11B will terminate treatment relatively early is not particularly restricted and can be any timing during the treatment of wafer WF in either reaction furnace 11A or 11B. However, when treatment is started in reaction furnaces 11A and 11B for the last or second-to-last wafer WF of the production batches present in loading ports 15X and 15Y, it is preferable to predict which reaction furnace 11A or 11B will terminate treatment relatively early. This is because the treatment is about to end, thus reducing the manufacturing error and increasing the accuracy of the prediction.

[0037] Additionally, the wafer manufacturing information (WF) used for prediction includes the processing time for wafer WF, the processing sequence of the wafer, and the number of wafer WF contained in the production batch. For example, using manufacturing information where the production batch contains five wafer WF and the processing time for a single wafer WF is n hours, the last or penultimate wafer WF can be identified, and the time to complete the processing of the last wafer WF can be calculated.

[0038] When the treatment of a production batch in one of the reaction furnaces 11A or 11B is completed, it is preferable to start the treatment of another production batch without an interval in either reaction furnace 11A or 11B if another production batch to be treated in either reaction furnace 11A or 11B is present in one of the loading ports. This allows the time loss to be as close to zero as possible.

[0039] The central drawing of Fig. Figure 3 is a time diagram showing an example of the handling of wafer WF from the same production batch using vapor phase growth device 1. Fig. 2 shows, and Fig. 4A to Fig. 4D are time diagrams that represent the procedures of the example in Fig. 3 illustrate. Additionally, Fig. 5 A flowchart illustrating the steps of handling the production batch and the wafer WF by the vapor phase growth device 1 of the present embodiment. With reference to the central drawing of Fig. 3 and Fig. 4A to Fig. 4D describes the steps of handling the production batch and the wafer WF of the present embodiment according to the flowchart in Fig. 5 described.

[0040] A specific example of the production plan for the manufacturing batch is described below, as shown in the drawing above. Fig. Figure 3 illustrates that, however, when a command is issued from the integrated control 16 of the vapor phase growth device 1 via the production management device 2 to the transport device 3, regardless of the sequence of the production plan indicated by the pi number, the wafer storage container 15 of the desired production batch can be transported to the desired loading ports 15X, 15Y and 15Z.

[0041] At the start of a day's operation, there is initially no wafer WF in either reaction furnace 11A or 11B, and no wafer storage bin 15 is present in any of the loading ports 15X, 15Y, or 15Z. Therefore, in step S1, the wafer storage bin 15 for the production batch to be processed in each of reaction furnaces 11A and 11B is transferred to loading ports 15X and 15Y, leaving the remaining loading port 15Z empty. This state at time T0 is shown in a central drawing of 4A. The first production batch is transferred to loading port 15X, the second production batch is transferred to loading port 15Y, and loading port 15Z is left empty.

[0042] In the next step S2, the wafers WF of the first production batch, which are transported into loading port 15X, are individually transported to the reaction furnace 11A to start the treatment, while the wafers WF of the second production batch, which are transported into loading port 15Y, are individually transported to the reaction furnace 11B to start the treatment.

[0043] In step S3, it is determined whether the production batch being transported to both loading port 15X and loading port 15Y is complete or nearly complete (for example, the penultimate wafer). This determination can be made from the number of wafers WF (product information) contained in the production batch, as output by the production management device 2, and the current number of processed wafers, calculated by the integrated control 16 of the vapor phase growth device 1. In step S3, if the production batch being transported to both loading port 15X and loading port 15Y is not complete or nearly complete, the process returns to step S2 and continues the treatment in reaction furnaces 11A and 11B.

[0044] In step S3, when one of the production batches being transported to loading port 15X and loading port 15Y is finished or nearly finished, the process continues to step S4 to determine which of the two reaction furnaces, 11A and 11B, currently undergoing treatment will terminate the treatment early. This state is shown at time T1 in the central drawing of 4A. Since the second production batch being transported to loading port 15Y is nearing completion, at this time T1, the time to complete the treatment in reaction furnace 11B is calculated based on the number of wafers WF (one wafer for the last wafer WF, two wafers for the second-to-last wafer) that have not yet completed the treatment in reaction furnace 11B and the processing time in reaction furnace 11B.Simultaneously, at time T1, the time to complete the treatment in reaction furnace 11A is calculated based on the number of wafers WF that have not yet completed the treatment in reaction furnace 11A and the processing time per wafer. These times are compared to determine which of the two reaction furnaces, 11A and 11B, completes the treatment first. In this example, reaction furnace 11B completes the treatment first.

[0045] In step S5, it is determined whether there is any empty space in loading ports 15X, 15Y, and 15Z. If no empty space is found, the step proceeds to step S7, but in this case, it proceeds to step S6 since loading port 15Z is empty. In step S6, the wafer storage bin 15 of the production batch to be processed in reaction furnace 11B, which is assumed to complete the processing relatively early, is selected from the production batch of the production plan and transported to loading port 15Z. As shown in the central drawing of Fig. As shown in 4A, the fourth production batch is transported to loading port 15Z. Although the sequence in the production plan is based on the third, since the third production batch is a batch to be processed in reaction furnace 11A, it is not transported to loading port 15Z.

[0046] In the subsequent step S7, it is determined whether all wafers (WF) of the production batch being transported into each loading port 15X, 15Y, and 15Z have completed the treatment. Production batches in which all wafers (WF) have not completed the treatment return to step S2 and continue the treatment. In step S7, the production batch in which all wafers (WF) have completed the treatment moves to step S8. In step S8, the wafer storage container 15 of the finished production batch is transported out of its loading port by the transport device 3. For example, as shown in the central drawing of Fig. As shown in Figure 4A, once all wafers WF of the second production batch, which are transported into loading port 15Y, have completed the treatment, the transport device removes this wafer storage container 15 from loading port 15Y. This leaves loading port 15Y empty, and simultaneously, reaction furnace 11B also moves to a state without wafers WF.

[0047] In the subsequent step S9, it is determined whether all production batches in the production plan have been transported. This routine ends when it is complete, but if the production batches have not been transported, the process returns to step S2 and continues the treatment in reaction furnaces 11A and 11B. Here, as shown in the central drawing of Fig. As shown in Figure 4A, reaction furnace 11B is also in a state without wafer WF. Furthermore, the fourth production batch to be processed in reaction furnace 11B is transported into loading port 15Z. Therefore, the wafers WF of this fourth production batch are transported individually to reaction furnace 11B, and the treatment is started. This allows reaction furnace 11B to operate continuously from the end of the treatment for the second production batch until the start of the treatment for the fourth production batch, and the time loss can be kept as close to zero as possible.

[0048] When the treatment for the fourth production batch is started in reaction furnace 11B, the subsequent step S3 determines again whether the production batch being transported to both loading port 15X and loading port 15Y is complete or nearly complete (for example, the penultimate wafer). Then, in step S3, if either of the production batches being transported to loading port 15X or loading port 15Z is complete or nearly complete, the process proceeds to step S4 to determine which of the two reaction furnaces, 11A and 11B, currently undergoing treatment will terminate the treatment early. This state is shown at time T2 in the lower diagram. Fig. Figure 4A shows that the first production batch being transported to loading port 15X is nearing completion. Therefore, at time T2, the time to complete treatment in reaction furnace 11A is calculated based on the number of wafers WF that have not yet completed treatment in reaction furnace 11A and the processing time in reaction furnace 11A. Simultaneously, at time T2, the time to complete treatment in reaction furnace 11B is calculated based on the number of wafers WF that have not yet completed treatment in reaction furnace 11B and the processing time per wafer. These times are compared to determine which of the two reaction furnaces, 11A and 11B, completes treatment first. In this example, reaction furnace 11A completes treatment first.

[0049] In step S5, it is determined whether empty space exists in loading ports 15X, 15Y, and 15Z. If so, the step proceeds to step S6, as loading port 15Y is empty. In step S6, the wafer storage bin 15 of the production batch to be processed in reaction furnace 11A, which is assumed to have completed processing relatively early, is selected from the production batch of the production plan and transported to loading port 15Y. Since the third production batch is one that should be processed in reaction furnace 11A, the third production batch is transported to loading port 15Y, as shown in the lower drawing. Fig. 4A shown.

[0050] In the subsequent step S7, it is determined whether all wafers WF of the production batch, which are transported into each loading port 15X, 15Y and 15Z, have completed the treatment, and as shown in the lower drawing of Fig. As shown in Figure 4A, after all wafers (WF) of the first production batch, which are transported into loading port 15X, have completed their treatment, the transport device removes this wafer storage container 15 from the transport device 15X. This leaves loading port 15X empty, and simultaneously, the reaction furnace 11A also moves to a state without wafers (WF). Since there are no wafers (WF) in the reaction furnace 11A, and furthermore, the third production batch to be treated in the reaction furnace 11A is being transported into loading port 15Y, the wafers (WF) of this third production batch are transported individually to the reaction furnace 11A, and the treatment is started. This allows the reaction furnace 11A to operate continuously from the end of the treatment for the first production batch until the start of the treatment for the third production batch, and the time loss can be kept as close to zero as possible.

[0051] At each of the points in time T3 to T8, which are in the Fig. 4B to Fig. As illustrated in Figure 4D, a similar determination is made in step S4 to assume which of the two reaction furnaces, 11A and 11B, will terminate the treatment early. Then, the wafer storage bin 15 of the production batch to be treated in reaction furnace 11A or 11B, which is assumed to terminate the treatment early, is transferred to the empty loading ports 15X, 15Y, or 15Z. This allows reaction furnaces 11A and 11B to operate continuously from the end of treatment for one production batch until the start of treatment for another, and the time loss can be kept as close to zero as possible.

[0052] Fig. Figure 6 is a block diagram illustrating an embodiment of a semiconductor manufacturing plant in which several semiconductor manufacturing devices are arranged. Fig. 1 are installed, and in particular the block diagram illustrates a semiconductor fabrication plant with three vapor phase growth devices 101, 102 and 103. Each of the vapor phase growth devices 101, 102 and 103 has the same configuration as the vapor phase growth device 1, which is shown in the Fig. 1 and Fig. Figure 2 illustrates this, and the vapor phase growth device 101 produces product specification A in one reaction furnace and produces product specification B in another reaction furnace; the vapor phase growth device 102 produces product specification B in both reaction furnaces; and the vapor phase growth device 103 produces product specification B in one reaction furnace and produces product specification C in another reaction furnace.

[0053] The integrated controller 16 is a computer that integrates the control of three vapor-phase growth devices 101, 102, and 103, respectively, and controls the operation of each of these devices. Additionally, the production information for each reaction furnace, sent by each of the three vapor-phase growth devices 101, 102, and 103, is aggregated to predict which reaction furnace will terminate the treatment relatively early. For example, since product specification B can be processed by the three vapor-phase growth devices 101, 102, and 103, the production information of the reaction furnace processing product specification B can be aggregated to predict which reaction furnace will terminate the treatment relatively early, thus reducing the waiting time for the product.

[0054] The production management device 2 is a device that integrally manages all manufacturing devices of the production process of the semiconductor manufacturing plant, including three vapor phase growth devices 101, 102 and 103, and instructs the three vapor phase growth devices 101, 102 and 103, based on the prediction from the integrated control 16, to transport the production batch to be treated in the reaction furnace that was predicted to finish treatment relatively early, to the loader-unloader that is currently empty, before the treatment in the reaction furnaces for the three vapor phase growth devices 101, 102 and 103 respectively is completed.

[0055] The transport device 3 is a system that transports the wafer WF, which has completed the treatment in the previous step, in a state in which it is stored in the wafer storage container 15, to the next step, and based on the command from the production management device 2, the wafer storage container 15 of the predetermined production batch is transported in the previous step into one of the loading ports 1x, 1y, 1z, 2x, 2y, 2z, 3x, 3y and 3z of the three vapor phase growth devices 101, 102 and 103.

[0056] The vapor phase growth apparatus 1 corresponds to the semiconductor manufacturing apparatus of the present invention, the reaction furnaces 11A and 11B correspond to the processor of the present invention, the loading ports 15X, 15Y and 15Z correspond to the loader-discharger of the present invention and the wafer WF corresponds to the object to be treated of the present invention. DESCRIPTION OF THE REFERENCE NUMBERS 1 Vapor phase growth device 11A, 11B Reaction furnace (processor) 111A, 111B Reaction chamber 112A, 112B Susceptor 113A, 113B Gas supply device 114A, 114B Shut-off valve 12 Wafer transfer chamber 121 First robot 122 First robot control 123 First Blade 13A, 13B Loading lock chamber 131A, 131B First Door 132A, 132B Second Door 14 Factory interface 141 Second robot 142 Second robot control 143 Second Blade 15 wafer storage containers 15X, 15Y, 15Z Charging port (charger-discharger) 16 Integrated control 2 Production management device 3 Transport device WF Wafer (object to be treated) QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] JP 2010-251507

[0003]

Claims

[1] Semiconductor manufacturing apparatus comprising a plurality of processors and a number of loader-unloaders exceeding the latter, wherein a processor for treatment transports an object to be treated, which is determined for each production batch, into the loader-unloader in a unit of the production batch, wherein each object to be treated is transported to the processor in a single unit and treated and is then transported to the loader-unloader, wherein the production batch of the object to be treated is transported by each processor into each loader-unloader, while the remaining loader-unloader is left empty, after which each of the processors begins the treatment; While the object to be processed is handled by each processor, the processor that terminates the processing relatively early is predicted for each processor based on manufacturing information of the object to be processed; and Before each processor finishes its treatment, the production batch to be treated by the processor that was predicted to finish treatment relatively early is transported to the loader-unloader, which is currently empty. [2] Semiconductor manufacturing apparatus according to claim 1, wherein, when the processor starts the treatment for a last or penultimate object to be treated of the manufacturing batch present in the loader-unloader, the processor that finishes the treatment relatively early is predicted. [3] Semiconductor manufacturing apparatus according to one of claims 1 and 2, wherein the manufacturing information of the object to be processed includes a processing time for the object to be processed and the number of objects to be processed that are to be included in a processing sequence or the manufacturing batch. [4] Semiconductor manufacturing apparatus according to one of claims 1 and 2, wherein, when the treatment of one production batch in one processor is completed, the treatment of another production batch by that processor is started when another production batch to be treated by that processor is present in one of the loader-dischargers. [5] Semiconductor manufacturing plant in which several semiconductor manufacturing devices according to claim 1 are installed. [6] Semiconductor manufacturing plant according to claim 5, comprising the following: an integrated control system that aggregates the processor manufacturing information sent by the multiple semiconductor manufacturing devices and predicts which processor will complete the processing relatively early; a production management device that, based on the prediction of the integrated control, instructs the multiple semiconductor manufacturing devices to transfer the production batch to be processed by the processor that was predicted to finish processing relatively early into the loader-unloader, which is currently empty, before the processing by each processor is completed; and a transport device that transports the production batch to the multiple semiconductor manufacturing devices based on the instruction from the production management device. [7] Semiconductor manufacturing process which uses a semiconductor manufacturing apparatus which is equipped with a plurality of processors and a number of loaders and unloaders which exceeds the plurality of processors, which transports by the processor to treat an object to be treated, which is determined for each production batch, into the loader and unloader in one unit of the production batch, then transports to the processor and treats each object to be treated in a single unit and then transports each object to be treated to the loader and unloader, wherein the process comprises: Treatment is initiated by each processor after the manufacturing batch of the object to be treated has been transported by each processor into each loader-unloader, while the remaining loader-unloader is left empty; Predictions of the processor that terminates processing relatively early, while the object to be processed is processed by each processor, based on the manufacturing information of the object to be processed for each processor; and Transport, before the end of the treatment by each processor, the production batch to be treated by the processor that was predicted to end the treatment relatively early, into the loader-unloader which is currently empty. [8] Semiconductor manufacturing process according to claim 7, wherein, when the processor starts the treatment for a last or penultimate object to be treated of the production batch present in the loader-unloader, the processor that finishes the treatment relatively early is predicted. [9] Semiconductor manufacturing process according to one of claims 7 and 8, wherein the manufacturing information of the object to be processed includes a processing time for the object to be processed and the number of objects to be processed that are to be included in a processing sequence or the manufacturing batch. [10] Semiconductor manufacturing process according to one of claims 7 and 8, wherein, when the treatment of one production batch in a processor is completed, the treatment of another production batch by that processor is started when another production batch to be treated by that processor is present in one of the loader-dischargers.

Images