Temperature estimation device, plasma processing system, temperature estimation method, and temperature estimation program

Abstract

To provide a temperature estimation device, plasma processing system, temperature estimation method, and temperature estimation program that estimate temperature corresponding to changes in the state within a processing space where plasma processing is performed.SOLUTION: The temperature estimation device has an estimation unit that sequentially inputs time-series process data related to the state within a processing space into a pre-generated time-series model generated by determining parameters of a time-series model on the basis of the error between temperature data sequentially output from the time-series model and temperature data measured at each time point within the processing space and sequentially inputs new time-series process data elated to the state within a processing space, thereby sequentially estimates temperature data.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a temperature estimation apparatus, a plasma processing system, a temperature estimation method, and a temperature estimation program. [Background technology]

[0002] In semiconductor manufacturing processes, it is important to maintain the temperature within a predetermined range in order to maintain the quality of wafers being plasma-processed. However, in semiconductor manufacturing processes, it is difficult to directly measure the temperature within the processing space by permanently installing a temperature sensor within the processing space. Therefore, the temperature within the processing space must be determined indirectly, for example, by estimation.

[0003] For example, Patent Document 1 below proposes a method for estimating the temperature in a processing space using process conditions during plasma processing. According to this estimation method, the temperature in the processing space during plasma processing can be estimated by acquiring the process conditions before the start of plasma processing.

[0004] However, even under the same process conditions, if the conditions in the processing space change (for example, if the temperature rises / falls at the start of plasma processing, or if the temperature of parts in the processing space rises during plasma processing), the accuracy of the temperature estimation decreases. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Publication No. 2019-212852 Summary of the Invention [Problem to be solved by the invention]

[0006] The present disclosure provides a temperature estimation device, a plasma processing system, a temperature estimation method, and a temperature estimation program for estimating a temperature according to a change in a state in a processing space where plasma processing is performed. [Means for solving the problem]

[0007] A temperature estimation device according to an aspect of the present disclosure has, for example, the following configuration: The apparatus has an estimation unit that sequentially inputs time-series process data related to the state in the processing space, and sequentially estimates temperature data by sequentially inputting new time-series process data related to the state in the processing space into a generated time-series model that is generated by determining parameters of the time-series model based on an error between temperature data sequentially output from the time-series model and data at each time of the time-series temperature data measured in the processing space. [Effects of the Invention]

[0008] According to the present disclosure, it is possible to provide a temperature estimation device, a plasma processing system, a temperature estimation method, and a temperature estimation program that estimate a temperature according to a change in the state in a processing space where plasma processing is performed. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 illustrates an example of a system configuration of a plasma processing system to which a temperature estimation apparatus in a model generation phase is applied. [Figure 2] FIG. 1 illustrates an example of a system configuration of a plasma processing system to which a temperature estimation device in an estimation phase is applied. [Figure 3] FIG. 2 is a diagram illustrating an example of a hardware configuration of a temperature estimation device. [Figure 4] FIG. 2 is a diagram showing an example of model generation data (a group of process data for model generation and temperature data for model generation). [Figure 5] FIG. 10 is a diagram illustrating an example of actual measurement values ​​of process data and temperature data. [Figure 6]10 is a flowchart showing the flow of a model generation process. [Figure 7] 10A and 10B are diagrams illustrating a specific example of a functional configuration and processing of an estimation unit. [Figure 8] 3A and 3B are diagrams illustrating a specific example of a functional configuration and processing of a determination unit. [Figure 9] FIG. 2 is a diagram illustrating a specific example of a functional configuration and processing of a control unit. [Figure 10] 10 is a flowchart showing the flow of an estimation process. DETAILED DESCRIPTION OF THE INVENTION

[0010] Hereinafter, each embodiment will be described with reference to the accompanying drawings. In this specification and drawings, components having substantially the same functional configurations are designated by the same reference numerals, and redundant description will be omitted.

[0011] [First embodiment] <System configuration of plasma processing system> First, a system configuration of a plasma processing system to which a temperature estimation apparatus according to a first embodiment is applied will be described. In the first embodiment, the temperature estimation apparatus has a function executed in a model generation phase and a function executed in an estimation phase. Therefore, the plasma processing system to which the temperature estimation apparatus is applied, which executes each of the functions, will be described separately below.

[0012] (1) System configuration of a plasma processing system to which a temperature estimation device in the model generation phase is applied 1 is a diagram showing an example of a system configuration of a plasma processing system to which a temperature estimation device in a model generation phase is applied. As shown in FIG. 1, in the model generation phase, plasma processing system 100 includes a semiconductor manufacturing process, time-series data acquisition devices 140_1 to 140_n, and a temperature estimation device 160.

[0013] In a semiconductor manufacturing process, plasma processing is performed on an object (unprocessed wafer 110) in a chamber 120, which is a processing space where plasma processing is performed, to generate a resultant object (processed wafer 130). Note that the unprocessed wafer 110 here refers to the wafer (substrate) before plasma processing is performed in the chamber 120, and the processed wafer 130 refers to the wafer (substrate) after plasma processing has been performed in the chamber 120.

[0014] In the model generation phase, a temperature sensor is installed in the chamber 120. The temperature sensor measures time-series data (temperature data for model generation) before plasma processing of the unprocessed wafer 110 is started in the chamber 120 or while the plasma processing is being performed. The example in FIG. 1 shows a state in which a temperature sensor 151 is installed at the position of the upper electrode in the chamber 120, and a temperature sensor 152 is installed at the position of the sidewall in the chamber 120. Note that the installation positions of the temperature sensors 151 and 152 in the chamber 120 are just an example, and they may be installed at positions different from those shown in FIG. 1. Furthermore, the number of temperature sensors installed in the chamber 120 is not limited to two.

[0015] The time-series data (temperature data for generating a model) measured by the temperature sensors 151 and 152 is stored in a model generation data storage unit 163 of the temperature estimation device 160.

[0016] The time-series data acquiring devices 140_1 to 140_n each measure a group of time-series data (a group of process data for generating a model) before plasma processing of the unprocessed wafer 110 is started in the chamber 120 or while the plasma processing is being performed. The time-series data acquiring devices 140_1 to 140_n are configured, for example, with a plurality of sensors (for example, n sensors), and each time-series data acquiring device measures process data for different types of measurement items. The number of measurement items measured by each of the time-series data acquiring devices 140_1 to 140_n may be one or more.

[0017] The time series data group (process data group for model generation) measured by the time series data acquiring devices 140_1 to 140_n is stored in the model generation data storage unit 163 in association with the time series data (temperature data for model generation) in the model generation phase.

[0018] A temperature estimation program is installed in the temperature estimation device 160. In the model generation phase, the temperature estimation device 160 functions as a model generation unit 161 by executing the program.

[0019] The model generation unit 161 performs a model generation process for a time series model using the model generation data (a group of process data for model generation, temperature data for model generation) stored in the model generation data storage unit 163, and provides a generated time series model.

[0020] (2) System configuration of a plasma processing system to which a temperature estimation device in the estimation phase is applied 2 is a diagram showing an example of the system configuration of a plasma processing system to which a temperature estimation device in the estimation phase is applied. The differences from FIG. 1 are that a temperature sensor is not installed in chamber 120 and that a function different from that in the model generation phase is performed in temperature estimation device 160.

[0021] By executing the temperature estimation program in the estimation phase, the temperature estimation device 160 functions as an estimation unit 200, a determination unit 210, and a control unit 220.

[0022] The estimation unit 200 has a generated time series model provided by the model generation unit 161 in the model generation phase. The estimation unit 200 sequentially inputs data for each time of the time series data group (process data group for estimation) measured by the time series data acquisition devices 140_1 to 140_n and stored in the time series data group storage unit 230 to the generated time series model. As a result, the estimation unit 200 acquires estimated temperature data sequentially output from the generated time series model. In addition, the estimation unit 200 notifies either or both of the determination unit 210 and the control unit 220 of the acquired estimated temperature data.

[0023] The determining unit 210 constantly monitors the current estimated temperature data in the chamber 120 notified by the estimating unit 200, and issues an alarm in real time if the temperature exceeds a predetermined temperature range. This allows the user of the plasma processing system 100 to know in real time that the temperature in the chamber 120 during plasma processing has exceeded the predetermined temperature range.

[0024] Furthermore, it is assumed that estimated temperature data is notified from the estimation unit 200 based on, for example, a time-series data group during plasma processing of a specific processed wafer 130, among past time-series data groups (process data groups for estimation) stored in the time-series data group storage unit 230. In this case, the determination unit 210 visualizes the time range during which the plasma processing was performed beyond the predetermined temperature range. This allows the user of the plasma processing system 100 to retroactively determine whether the temperature during plasma processing of the specific processed wafer 130 exceeded the predetermined temperature range.

[0025] The control unit 220 controls the semiconductor manufacturing process in real time (for example, by appropriately setting the amount of power of the RF power supply or the processing time of the plasma processing) so that the current estimated temperature data in the chamber 120 notified by the estimation unit 200 reaches the target temperature. This makes it possible to adjust the temperature at the start of plasma processing of the unprocessed wafer 110 to the target temperature, for example.

[0026] The control unit 220 also determines whether the process conditions are appropriate based on estimated temperature data in the chamber 120 (e.g., estimated temperature data during plasma processing on the immediately preceding unprocessed wafer 110) notified by the estimation unit 200. Furthermore, the control unit 220 optimizes the process conditions when performing plasma processing on the next unprocessed wafer 110 (e.g., appropriately sets the power amount of the RF power supply or the processing time for plasma processing). This allows, for example, the temperature when performing plasma processing on the next unprocessed wafer 110 to be kept within a predetermined temperature range.

[0027] <Hardware configuration of the temperature estimation device> Next, the hardware configuration of the temperature estimation device 160 will be described. Fig. 3 is a diagram showing an example of the hardware configuration of the temperature estimation device. As shown in Fig. 3, the temperature estimation device 160 has a CPU (Central Processing Unit) 301, a ROM (Read Only Memory) 302, and a RAM (Random Access Memory) 303. The temperature estimation device 160 also has a GPU (Graphics Processing Unit) 304. Processors (processing circuits, processing circuitry) such as the CPU 301 and GPU 304 and memories such as the ROM 302 and RAM 303 form a so-called computer.

[0028] Furthermore, the temperature estimation device 160 has an auxiliary storage device 305, a display device 306, an operation device 307, an I / F (Interface) device 308, and a drive device 309. The hardware components of the temperature estimation device 160 are connected to each other via a bus 310.

[0029] The CPU 301 is a computing device that executes various programs (for example, a temperature estimation program) installed in the auxiliary storage device 305 .

[0030] The ROM 302 is a non-volatile memory and functions as a main storage device. The ROM 302 stores various programs, data, etc. required for the CPU 301 to execute various programs installed in the auxiliary storage device 305. Specifically, the ROM 302 stores boot programs such as a BIOS (Basic Input / Output System) and an EFI (Extensible Firmware Interface).

[0031] The RAM 303 is a volatile memory such as a dynamic random access memory (DRAM) or a static random access memory (SRAM), and functions as a main storage device. The RAM 303 provides a working area into which various programs installed in the auxiliary storage device 305 are expanded when executed by the CPU 301.

[0032] The GPU 304 is a computing device for image processing, and in this embodiment, performs high-speed calculations using parallel processing on a time-series data group when the temperature estimation program is executed by the CPU 301. The GPU 304 is equipped with an internal memory (GPU memory) and temporarily stores information required when performing parallel processing on a time-series data group.

[0033] The auxiliary storage device 305 stores various programs and various data used when the various programs are executed by the CPU 301. For example, the model generation data storage unit 163 and the time-series data group storage unit 230 are realized in the auxiliary storage device 305.

[0034] The display device 306 is, for example, a display device that displays an alarm to a user of the plasma processing system 100. The operation device 307 is, for example, an input device used by a user of the plasma processing system 100 or an administrator of the temperature estimation device 160 when inputting various instructions to the temperature estimation device 160. The I / F device 308 is a connection device that connects to a network (not shown) and performs communication.

[0035] The drive device 309 is a device for loading a recording medium 320. The recording medium 320 here includes media that record information optically, electrically, or magnetically, such as a CD-ROM, a flexible disk, a magneto-optical disk, etc. The recording medium 320 may also include semiconductor memory that records information electrically, such as a ROM, a flash memory, etc.

[0036] The various programs to be installed in the auxiliary storage device 305 are installed, for example, by setting the distributed recording medium 320 in the drive device 309 and reading the various programs recorded on the recording medium 320 by the drive device 309. Alternatively, the various programs to be installed in the auxiliary storage device 305 may be installed by being downloaded via a network (not shown).

[0037] <Examples of data for model generation> Next, a specific example of model generation data (a group of process data for model generation and temperature data for model generation) stored in the model generation data storage unit 163 will be described. Fig. 4 is a diagram showing an example of model generation data (a group of process data for model generation and temperature data for model generation). In the example of Fig. 4, for the sake of simplicity, each of the time-series data acquiring devices 140_1 to 140_n measures one-dimensional data, but one time-series data acquiring device may measure two-dimensional data (a data set of multiple types of one-dimensional data).

[0038] FIG. 4(a) schematically shows a time-series data group (a process data group for generating a model) made up of time-series data measured in the same time range by the time-series data acquiring devices 140_1 to 140_n.

[0039] In FIG. 4(a), for example, time series data 1 is a schematic representation of predetermined gas flow rate data (for example, data measured by a sensor that measures the gas flow rate of a gas supplied into chamber 120). Note that the gas referred to here includes all types of gas (process gas and inert gas) supplied into chamber 120. Furthermore, the sensors that measure gas flow rates include sensors provided in the gas lines of each type of gas (however, for convenience, only time series data 1 is shown here).

[0040] 4(a), for example, time series data 2 is a schematic representation of pressure data within the chamber 120 (for example, data measured by a sensor that measures the pressure within the chamber 120). For example, time series data 3 is a schematic representation of RF power supply data (for example, data measured by a sensor that measures the amount of power of an RF power supply for plasma processing).

[0041] 4(a), for example, time series data n-1 is a schematic representation of impedance data (for example, data measured by a sensor for measuring chamber impedance (matching position, current, voltage, etc.)). Furthermore, for example, time series data n is a schematic representation of heater power data (for example, data measured by a sensor for measuring the amount of power of a heater that heats the inside of chamber 120).

[0042] 4(b) is a schematic diagram showing time-series data (temperature data for generating a model) measured in a corresponding time range by a temperature sensor 151 attached to the position of the upper electrode, and FIG. 4(b) is a schematic diagram showing time-series data (temperature data for generating a model) measured in a corresponding time range by a temperature sensor 152 attached to the position of the inner wall of the chamber 120.

[0043] <Relationship between process data groups and temperature data> Next, the relationship between the time series data group (process data group for generating a model) measured by the time series data acquiring devices 140_1 to 140_n and the time series data (temperature data for generating a model) measured by the temperature sensors 151 and 152 will be described.

[0044] 5A and 5B are diagrams showing examples of measured process data and temperature data. Of these, Fig. 5A shows measured RF power data, with the horizontal axis representing time and the vertical axis representing power level. Fig. 5B shows measured temperature data measured by a temperature sensor 151 attached to the upper electrode, with the horizontal axis representing time and the vertical axis representing temperature.

[0045] 5(a) and 5(b), the temperature data inside the chamber 120 fluctuates with a certain time delay relative to fluctuations in the process data. This is because it takes a certain amount of time for the thermally conductive materials constituting each part inside the chamber 120 to conduct heat. Also, it takes a certain amount of time for the temperature of each part to stabilize after heat is conducted to each part inside the chamber 120.

[0046] In view of such temperature characteristics within the chamber 120, the temperature estimation device 160 according to this embodiment performs a model generation process for a time series model, which is a model capable of expressing time delays, and estimates temperature data using the generated time series model provided.

[0047] <Overview of the model generation section> Next, an overview of the model generation unit 161 of the temperature estimation device 160 will be described. The model generation unit 161 has a time series model such as an ARX model. The model generation unit 161 performs a batch-type model generation process on the time series model using model generation data (a group of process data for model generation, temperature data for model generation) stored in a model generation data storage unit 163.

[0048] Specifically, the model generation unit 161 reads out a group of process data for model generation from the model generation data storage unit 163 and sequentially inputs the data to the time series model. The model generation unit 161 also compares the output results sequentially output from the time series model with the corresponding temperature data for model generation (for example, temperature data measured by the temperature sensor 151 attached to the position of the upper electrode) and calculates an error for each.

[0049] Furthermore, the model generating unit 161 determines the model parameters of the time series model so that the sum of squares of the calculated errors is minimized.

[0050] The generated time series model with the determined model parameters can be expressed as, for example,

[0051]

number

[0052] As is clear from the above equation, in the time series model, the time range m b and a set of process data from a given time range m a The current temperature data is calculated by weighting and adding the temperature data of the previous temperature data and the previous temperature data using the determined model parameters.

[0053] As a result, the temperature estimation device 160 can estimate temperature data according to fluctuations in process data using the generated time series model, even if the state inside the chamber 120 has changed (or has changed).

[0054] Here, the case where the model parameters of the time series model are determined based on the temperature data measured by the temperature sensor 151 attached to the position of the upper electrode has been described. However, the model parameters of the time series model can also be determined by performing the same process on the temperature data measured by the temperature sensor 152 attached to the position of the side wall inside the chamber 120.

[0055] However, since different model parameters are obtained when the temperature data measured by the temperature sensor 151 is used and when the temperature data measured by the temperature sensor 152 is used, the model generation unit 161 manages each as a separate generated time series model.

[0056] In the above description, the predetermined time range m a , m b Although the above description does not mention a method for selecting the process data group to be input to the time series model, the order of the time series model is optimized by being adjusted as needed. Similarly, the above description does not mention a method for selecting the process data group to be input to the time series model from the process data group for model generation, but the process data group to be input to the time series model is also optimized by being adjusted as needed.

[0057] <Model generation process flow> Next, a flow of a model generation process for a plasma processing system in the model generation phase will be described with reference to a flowchart shown in FIG.

[0058] In step S601, the temperature estimation device 160 acquires model generation data and verification data. Specifically, the temperature estimation device 160 acquires, as model generation data, a group of process data and temperature data from the time-series data acquisition devices 140_1 to 140_n and the temperature sensor 151 (or 152). Similarly, the temperature estimation device 160 acquires, as verification data, a group of process data and temperature data from the time-series data acquisition devices 140_1 to 140_n and the temperature sensor 151 (or 152).

[0059] In step S602, the temperature estimation device 160 selects a group of process data necessary for estimating temperature data from the group of process data for model generation.

[0060] In step S603, the temperature estimation device 160 selects the order of the time series model that the model generation unit 161 has.

[0061] In step S604, the temperature estimation device 160 performs a model generation process on the time series model so as to minimize the sum of squares of the errors when the selected group of process data is input, and determines the model parameters of the time series model.

[0062] In step S605, the temperature estimation device 160 verifies the time series model (generated time series model) for which the model parameters have been determined, using the verification data.

[0063] As a result of the verification, it is determined whether the target estimation accuracy has been reached, and if it is determined that the target estimation accuracy has not been reached (No in step S605), the process returns to step S602. In this case, in step S602, the temperature estimation device 160 adds process data that contributes to improving the estimation accuracy to the existing selected process data group, or selects a process data group of a combination different from the previous combination.

[0064] On the other hand, if it is determined in step S605 that the target estimation accuracy has been reached (Yes in step S605), the model generation process ends.

[0065] <Specific examples of functional configuration and processing of the estimation unit> Next, a description will be given of a functional configuration of the estimation unit 200 and a specific example of processing by the estimation unit 200. Fig. 7 is a diagram showing a functional configuration of the estimation unit and a specific example of processing.

[0066] As shown in FIG. 7, the estimation unit 200 has a generated time series model 710 provided by the model generation unit 161.

[0067] The estimation unit 200 sequentially inputs the time series data groups (process data groups for estimation) acquired by the time series data acquiring devices 140_1 to 140_n to the generated time series model 710. The time series data groups (process data groups for estimation) input at this time are process data groups (time series process data groups related to the state of the chamber 120) selected when it is determined that the time series data groups have a predetermined estimation accuracy in the model generation process.

[0068] Before inputting the selected process data group, the estimation unit 200 executes, for example, a temperature initialization sequence as follows. The process for sufficiently stabilizing the temperature inside the chamber 120 is continued for a certain period of time or more, and at this time, the process data group (selected process data group) and temperature data are collected for a period of time or more that allows the model parameters to be identified. The collected process data group and temperature data are input as initial values ​​to the generated time series model 710.

[0069] In the example of FIG. 7, the current process data group is a time series data group at a time indicated by reference numeral 721, and shows how the process data group has been input.

[0070] As shown in FIG. 7, the generated time series model 710 includes a time range mb In addition, in the generated time series model 710, the process data set from the past m is used to calculate the estimated temperature data. a The estimated temperature data from the past is used to calculate the current estimated temperature data.

[0071] That is, in the generated time series model 710, The current process data set (721) is b a process data group from a predetermined time range past the current process data group; -The estimated temperature data of the current (code 721) is within the time range m a Previous estimated temperature data (estimated temperature data within a predetermined time range that is earlier than the estimated temperature data of the estimation target), The generated time series model 710 sequentially notifies the determination unit 210 and the control unit 220 of the calculated estimated temperature data of the current estimation target (reference numeral 721) based on the above.

[0072] In this way, the estimation unit 200 sequentially inputs new time-series process data sets related to the state inside the chamber 120 to the generated time-series model 710, thereby sequentially outputting estimated temperature data.

[0073] Although FIG. 7 illustrates the case of estimating estimated temperature data at the position of the upper electrode, when the generated time series model 710 is used to estimate estimated temperature data at the position of the side wall in the chamber 120, the estimated temperature data can be estimated sequentially.

[0074] <Specific examples of functional configuration and processing of the determination unit> Next, a description will be given of a functional configuration of the determination unit 210 and a specific example of processing by the determination unit 210. Fig. 8 is a diagram showing a functional configuration of the determination unit and a specific example of processing.

[0075] As shown in FIG. 8, the determination unit 210 includes a deviation amount calculation unit 810, a determination criterion setting unit 820, and a notification unit 830.

[0076] The deviation amount calculation unit 810 periodically acquires, for example, real-time estimated temperature data notified by the estimation unit 200. Furthermore, every time the deviation amount calculation unit 810 acquires real-time estimated temperature data, it determines whether the estimated temperature data exceeds the temperature range preset by the determination criterion setting unit 820.

[0077] Furthermore, when it is determined that the estimated temperature data exceeds a preset temperature range, deviation amount calculation unit 810 calculates the deviation amount and notifies notification unit 830. In Fig. 8, graph 840 shows real-time estimated temperature data notified to deviation amount calculation unit 810 by estimation unit 200. Note that graph 840 shows estimated temperature data when plasma processing is performed consecutively on a plurality of unprocessed wafers, with the horizontal axis representing the processing time of each unprocessed wafer and the vertical axis representing the estimated temperature data.

[0078] 8, graph 860 is an enlarged view of time range 841 of graph 840. In graph 860, dotted line 862 indicates the upper limit of the temperature range preset by judgment criterion setting unit 820, and dotted line 863 indicates the lower limit of the temperature range preset by judgment criterion setting unit 820.

[0079] According to graph 860, estimated temperature data 861 exceeds the predetermined temperature range (below the lower limit of dotted line 863 in the example of FIG. 8), so deviation amount calculation unit 810 determines that estimated temperature data 861 exceeds the preset temperature range and calculates the deviation amount. In this case, notification unit 830 issues an alarm to the user of plasma processing system 100 indicating that the predetermined temperature range has been exceeded.

[0080] 8, graph 850 is an enlarged view of time range 842 of graph 840. In graph 850, dotted line 852 indicates the upper limit of the temperature range preset by judgment criterion setting unit 820, and dotted line 853 indicates the lower limit of the temperature range preset by judgment criterion setting unit 820.

[0081] According to graph 850, estimated temperature data 851 is within the predetermined temperature range, and therefore deviation amount calculation unit 810 determines that estimated temperature data 851 does not exceed the preset temperature range. In this case, notification unit 830 does not issue an alarm to the user of plasma processing system 100.

[0082] In the above description, the determination unit 210 performs real-time determination processing on estimated temperature data notified by the estimation unit 200. However, the same applies to a case where the determination unit 210 performs determination processing on estimated temperature data notified by the estimation unit 200 based on a group of past time-series data stored in the model generation data storage unit 163.

[0083] In this case, however, the notification unit 830 notifies the user of the plasma processing system 100 by visualizing the time range during which the plasma processing was performed beyond the predetermined temperature range.

[0084] <Specific examples of functional configuration and processing of the control unit> Next, a description will be given of a functional configuration of the control unit 220 and a specific example of processing by the control unit 220. Fig. 9 is a diagram showing a functional configuration of the control unit and a specific example of processing.

[0085] As shown in FIG. 9, the control unit 220 includes a temperature adjustment unit 910, a target temperature setting unit 920, a recipe adjustment unit 930, and the like.

[0086] The temperature adjusting unit 910 generates control data based on, for example, real-time estimated temperature data notified by the estimating unit 200 during the seasoning process so that the current estimated temperature data in the chamber 120 reaches the target temperature. The target temperature is set by, for example, the target temperature setting unit 920.

[0087] Furthermore, the temperature adjusting unit 910 adjusts the temperature inside the chamber 120 by transmitting the generated control data to the semiconductor manufacturing process.

[0088] As a result, in the past, even if the temperature at the start of plasma processing had dropped to the temperature indicated by reference numeral 941, the estimated temperature data at that time could not be accurately estimated, and plasma processing would have been started as is. In contrast, by having the estimation unit 200 notify the current estimated temperature data in real time, the temperature adjustment unit 910 can adjust the temperature in the chamber 120 to reach the target temperature indicated by reference numeral 942, for example, during seasoning processing. As a result, the temperature change during plasma processing for the first unprocessed wafer is similar to the temperature change for the second and subsequent unprocessed wafers, and the quality of the first plasma-processed processed wafer can be maintained.

[0089] The recipe adjustment unit 930 determines the suitability of the process conditions based on, for example, estimated temperature data output during plasma processing (e.g., estimated temperature data output during plasma processing on the immediately preceding unprocessed wafer 110). Furthermore, the recipe adjustment unit 930 optimizes the process conditions for performing plasma processing on the next unprocessed wafer 110 based on the determination result of the suitability of the process conditions, and adjusts the recipe. Furthermore, the recipe adjustment unit 930 performs plasma processing using the adjusted recipe by transmitting the adjusted recipe to the semiconductor manufacturing process.

[0090] As a result, for example, while conventionally the same recipe was used even if the temperature inside the chamber 120 changed due to successive plasma processing operations, the recipe adjustment unit 930 can use a recipe that corresponds to the temperature change (see reference numerals 951 and 952). As a result, the temperature when performing plasma processing on the next unprocessed wafer 110 can be kept within a predetermined temperature range.

[0091] In the example of Figure 9, the function of adjusting the temperature inside the chamber 120 during seasoning processing and the function of adjusting the recipe during plasma processing are illustrated, but the functions that the control unit 220 adjusts based on estimated temperature data are not limited to these.

[0092] <Flow of estimation process> Next, the flow of the estimation process in the estimation phase of the plasma processing system will be described with reference to the flowchart of FIG.

[0093] In step S1001, the time series data acquiring devices 140_1 to 140_n start measuring a time series data group (a process data group for estimation) and store the measured time series data group in the time series data group storage unit 230.

[0094] In step S1002, the temperature estimation device 160 determines whether to perform real-time processing using the time-series data group. Note that the real-time processing refers to processing in which estimated temperature data is output in real time by inputting the time-series data group (process data group for estimation) before the plasma processing is started or during the plasma processing into the generated time-series model.

[0095] In step S1002, if it is determined that real-time processing is to be executed (if Yes in step S1002), the process proceeds to step S1011.

[0096] In step S1011, the temperature estimation device 160 inputs a group of process data (a group of time-series process data related to the chamber state) selected from the group of measured time-series data (a group of process data for estimation) into the generated time-series model. At the start of the estimation process, the temperature estimation device 160 inputs an initial value of the temperature data into the generated time-series model. This allows the temperature estimation device 160 to output estimated temperature data in real time.

[0097] In step S1012, the temperature estimation device 160 determines whether to execute the determination mode or the control mode. If it is determined in step S1012 that the determination mode is to be executed, the process proceeds to step S1013.

[0098] In step S1013, the temperature estimation device 160 performs real-time determination processing, and issues an alarm if the current estimated temperature data is outside a preset temperature range.

[0099] On the other hand, if it is determined in step S1012 that the control mode is to be executed, the process proceeds to step S1014.

[0100] In step S1014, the temperature estimation device 160 performs a temperature adjustment process to adjust the temperature in the chamber so that the current estimated temperature data reaches the target temperature. Alternatively, the temperature estimation device 160 determines whether the process conditions used to perform plasma processing on the immediately preceding unprocessed wafer are appropriate. Then, the temperature estimation device 160 optimizes the process conditions used to perform plasma processing on the next unprocessed wafer and adjusts the recipe.

[0101] On the other hand, if it is determined in step S1002 that real-time processing is not to be executed (No in step S1002), the process proceeds to step S1021.

[0102] In step S1021, the temperature estimation device 160 selects, from the time series data group (process data group for estimation) stored in the time series data group storage unit 230, For example, a time series data group during plasma processing of a specific processed wafer 130, -Selected process data group (time-series process data group related to the chamber state), The temperature estimation device 160 also inputs the read process data group into the generated time series model and outputs estimated temperature data.

[0103] In step S1022, temperature estimation device 160 performs a determination process, visualizes the time range in which plasma processing was performed beyond a predetermined temperature range from the estimated temperature data, and notifies the user of plasma processing system 100 of the visualized time range.

[0104] In step S1030, temperature estimation device 160 determines whether or not to end the estimation process. If it is determined in step S1030 that the estimation process is to be continued (No in step S1030), the process returns to step S1002.

[0105] On the other hand, if it is determined in step S1030 that the estimation process should be ended (Yes in step S1030), the estimation process is ended.

[0106] <Summary> As is clear from the above description, the temperature estimation device 160 according to the first embodiment: A model generation process is performed for a time series model that associates data for each time range of a time series process data group related to the state inside a chamber where plasma processing is performed with data for each time period of time series temperature data measured inside the chamber, thereby providing a generated time series model. New time-series process data related to the state inside the chamber are input sequentially into the provided generated time-series model, thereby sequentially estimating the temperature data.

[0107] In this way, in the first embodiment, in consideration of the unique circumstances of the semiconductor manufacturing process in which it is difficult to install a temperature sensor in the chamber to directly measure the temperature, the temperature in the chamber is indirectly estimated from a group of time-series process data. In this case, in the first embodiment, the time-series temperature data is estimated by using data for each time range of the group of time-series process data related to the state of the chamber.

[0108] As a result, according to the first embodiment, even if the state inside the chamber has changed (or has changed), it is possible to estimate temperature data according to fluctuations in process data.

[0109] In other words, according to the first embodiment, it is possible to provide a temperature estimation device, a plasma processing system, a temperature estimation method, and a temperature estimation program that estimate a temperature according to changes in the state within the processing space where plasma processing is performed.

[0110] [Second embodiment] In the first embodiment, the model generation unit 161 is described as performing the model generation process on an ARX model. However, the time series model on which the model generation unit 161 performs the model generation process is not limited to the ARX model. For example, any other model (e.g., an RNN (Recurrent Neural Network) model) may be used as long as it is a model capable of calculating the correlation of time series data.

[0111] In the case of an RNN model, the model generation unit 161 determines model parameters through a learning process and provides a trained RNN model. Specifically, in the case of an RNN model, the model generation unit 161 performs a learning process using a learning dataset in which a group of process data for model generation is used as input data and temperature data for model generation is used as correct data, and determines model parameters.

[0112] In other words, the model generation unit 161 functions as a learning unit, and performs learning processing using backpropagation or the like so that the output data when input data is input to the RNN model approaches the correct data, and provides a trained RNN model.

[0113] In the first embodiment, the recipe adjustment unit 930 generates an adjusted recipe based on estimated temperature data. However, the recipe adjustment unit 930 may, for example, analyze dynamic events in the chamber 120 based on the estimated temperature data and calculate characteristic values ​​for adjusting the recipe. The characteristic values ​​referred to here include, for example, a C / D (Critical Dimension) value and an E / R (Etching Rate) value.

[0114] In the first embodiment, the temperature adjustment unit 910 adjusts the temperature in the chamber 120 during the seasoning process. However, the temperature adjustment unit 910 may adjust the temperature in the chamber 120 not only during the seasoning process. For example, the temperature adjustment unit 910 may adjust the temperature during plasma processing of an unprocessed wafer. The plasma processing referred to here includes etching, film formation, ashing, and the like.

[0115] Furthermore, in the first embodiment, the semiconductor manufacturing process and the temperature estimation device 160 are configured as separate entities, but at least some of the functions of the temperature estimation device 160 may be configured to be realized in the semiconductor manufacturing process.

[0116] Furthermore, in the above first embodiment, the model generation phase and the estimation phase are described as being executed by the same temperature estimation device 160, but the temperature estimation device that executes the model generation phase and the temperature estimation device that executes the estimation phase may be configured separately.

[0117] In the first embodiment, the temperature estimation device 160 that executes the estimation phase has been described as functioning as the estimation unit 200, the determination unit 210, and the control unit 220. However, the temperature estimation device 160 that executes the estimation phase may be configured, for example, as being separated into the estimation unit 200, the determination unit 210, and the control unit 220. In other words, the temperature estimation device 160 may be configured as being separated into a temperature estimation device having the estimation unit 200, an alarm output device having the determination unit 210, and a temperature control device having the control unit 220.

[0118] The present invention is not limited to the configurations described in the above embodiments, but may be combined with other elements, etc. These aspects can be changed without departing from the spirit of the present invention, and can be appropriately determined depending on the application form. [Explanation of symbols]

[0119] 100: Plasma processing system 110: Unprocessed wafer 120: Chamber 130: Processed wafer 140_1~140_n: Time series data acquisition device 151: Temperature sensor 152: Temperature sensor 160: Temperature estimation device 161: Model generation unit 200: Estimation section 210: Judgment section 220: Control unit 710: Generated time series models 810: Deviation amount calculation unit 820: Judgment criteria setting section 830: Information Department 910: Temperature adjustment section 920:Target temperature setting section 930: Recipe Adjustment Department

Claims

1. an estimation unit that sequentially estimates temperature data by sequentially inputting new time-series process data related to the state in the processing space into a generated time-series model that is generated by sequentially inputting time-series process data related to the state in the processing space and determining parameters of the time-series model based on an error between temperature data sequentially output from the time-series model and data at each time of the time-series temperature data measured in the processing space; A temperature estimation device having:

2. The temperature estimation device according to claim 1 , wherein the new time-series process data relating to the state in the processing space is a plurality of time-series process data newly measured by a plurality of sensors, respectively.

3. 3. The temperature estimation device according to claim 2, wherein the plurality of sensors include at least one of a sensor for measuring a pressure within the processing space, a sensor for measuring a gas flow rate of a gas supplied into the processing space, a sensor for measuring an electric power amount of an RF power source for plasma processing, a sensor for measuring a chamber impedance, and a sensor for measuring an electric power amount of a heater for heating the processing space.

4. 3. The temperature estimation device according to claim 2, wherein the new time-series process data relating to the state within the processing space is a plurality of time-series process data newly measured before plasma processing is started or during plasma processing.

5. 3. The temperature estimation device according to claim 2, wherein the generated time series model performs weighted addition of process data in a predetermined time range older than current process data and temperature data estimated in a predetermined time range older than the temperature data of the estimation target, among the new time series process data, using parameters of the generated time series model, to output the temperature data of the estimation target.

6. 2. The temperature estimation device of claim 1, wherein the generated time series model is generated according to the measured position by determining parameters of the time series model based on an error between time series temperature data measured at a predetermined position in the processing space and data at each time.

7. The temperature estimation device according to any one of claims 1 to 6, a control device that sets a processing time of the plasma processing or an amount of power of an RF power source so that the temperature data sequentially estimated by the estimation unit reaches a target temperature before starting the plasma processing or during the plasma processing; A plasma processing system comprising:

8. 8. The plasma processing system according to claim 7, further comprising an alarm output device that outputs an alarm when the temperature data sequentially estimated by the estimation unit exceeds a predetermined temperature range.

9. a model generation unit that sequentially inputs time-series process data related to the state in the processing space, and determines parameters of the time-series model based on an error between temperature data sequentially output from the time-series model and data at each time of the time-series temperature data measured in the processing space, thereby providing a generated time-series model; A temperature estimation device having:

10. a step of sequentially inputting time-series process data relating to the state in the processing space, and sequentially estimating temperature data by sequentially inputting new time-series process data relating to the state in the processing space into a generated time-series model, the generated time-series model being generated based on an error between temperature data sequentially output from the time-series model and time-series temperature data measured in the processing space; A temperature estimation method having the following:

11. a step of sequentially inputting time-series process data relating to the state in the processing space, and sequentially estimating temperature data by sequentially inputting new time-series process data relating to the state in the processing space into a generated time-series model, the generated time-series model being generated based on an error between temperature data sequentially output from the time-series model and time-series temperature data measured in the processing space; A temperature estimation program for running on a computer.

12. a step of sequentially inputting time-series process data relating to the state in the processing space, and determining parameters of the time-series model based on an error between temperature data sequentially output from the time-series model and data at each time of the time-series temperature data measured in the processing space, thereby providing a generated time-series model; A temperature estimation method having the following:

13. a step of sequentially inputting time-series process data relating to the state in the processing space, and determining parameters of the time-series model based on an error between temperature data sequentially output from the time-series model and data at each time of the time-series temperature data measured in the processing space, thereby providing a generated time-series model; A temperature estimation program for running on a computer.

Images