Search device, plasma processing apparatus, and search method

Abstract

To further improve a prediction accuracy of a learning model in order to reduce the number of searches of a value of a parameter to be set to a plasma processing apparatus.SOLUTION: A search device 3 searches for a value of a first parameter to be set for a plasma processing apparatus 1 for processing a processing object into a target material by using a processing gas. A control part 39 selects a plurality of first combinations constructed by a value that may be taken by the first parameter. The control part 39 acquires a value of a second parameter indicating a state of a plasmatized processing gas to each combination of the first combination, and a value of a third parameter indicating an evaluation of the processing object after the processing. The control part 39 generates a learning model 82 in which the first parameter and the second parameter are defined as an explanatory variable, and the third parameter is defined as an objective variable. The control part 39 predicts the value of the first parameter so that each value of the third parameter becomes a predetermined value by using the learning model 82.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a detection device, a plasma processing device, and a detection method.

Background Art

[0002] As disclosed in Patent Document 1 (Japanese Patent Application Laid-Open No. 2019-165123), there is a technique of generating a learning model in which parameters set in a plasma processing device are explanatory variables and parameters indicating evaluation of a processed object after processing are objective variables, and using the learning model to search for values of parameters to be set in the plasma processing device for processing a processing object into a target object.

Summary of the Invention

Problems to be Solved by the Invention

[0003] In order to reduce the number of searches for values of parameters to be set in the plasma processing device, it is desirable to further improve the prediction accuracy of the learning model.

Means for Solving the Problems

[0004] The search device from the first perspective searches for the value of each of one or more first parameters, which sets the plasma processing device that processes the object to be processed into a target object using a process gas. The search device includes a control unit and a storage unit. The control unit selects a plurality of first combinations composed of values that each first parameter can take. When each combination of the first combination is set in the plasma processing device, the control unit acquires, for each combination of the first combination, the value of each of one or more second parameters indicating the state of the plasma-processed process gas and the value of each of one or more third parameters indicating the evaluation of the object to be processed after processing. The control unit stores, in the storage unit, learning data associating the combination of the first combination, the value of each of the acquired second parameters, and the value of each of the acquired third parameters. The control unit generates a learning model using the learning data. The learning model uses the first parameter and the second parameter as explanatory variables and the third parameter as the objective variable. The control unit predicts the value of each first parameter such that the value of each third parameter becomes a predetermined value using the learning model.

[0005] In the search device from the first perspective, the learning model uses, as explanatory variables, one or more second parameters indicating the state of the plasma-processed process gas in addition to one or more first parameters set in the plasma processing device. As a result, the search device can further improve the prediction accuracy of the learning model.

[0006] The search device from the second perspective is the search device from the first perspective, and the control unit selects a first combination having a smaller number of combinations than the second combination from among a plurality of second combinations composed of values that each first parameter can take by the experimental design method.

[0007] As a result, the search device can efficiently search for the value of the parameter set in the plasma processing device.

[0008] The exploration device from the third perspective is the exploration device from the first or second perspective, and the first parameter includes at least one of the power for plasma generation, the bias power, the pressure of the process gas, the flow rate of the process gas, and the type of the process gas.

[0009] The exploration device from the fourth perspective is the exploration device from any one of the first to third perspectives, and the second parameter includes a parameter that directly or indirectly represents at least one of the electron temperature, the electron density, the radical species, the radical density, the ion species, the ion density, and the ion energy.

[0010] The exploration device from the fifth perspective is the exploration device from the fourth perspective, and the parameter that indirectly represents the radical species and the parameter that indirectly represents the radical density include at least one of the emission wavelength of the plasma, the emission intensity of the emission wavelength of the plasma, the ratio of the emission intensities of the plurality of emission wavelengths of the plasma, the mass of the radicals in the plasma, the observed intensity of the mass of the radicals in the plasma, the ratio of the observed intensities of the masses of the plurality of radicals in the plasma, or a value obtained by calculation using at least one of them.

[0011] The exploration device from the sixth perspective is the exploration device from the fourth perspective, and the parameter that indirectly represents the ion species and the parameter that indirectly represents the ion density include at least one of the emission wavelength of the plasma, the emission intensity of the emission wavelength of the plasma, the ratio of the emission intensities of the plurality of emission wavelengths of the plasma, the mass of the ions, the observed intensity of the mass of the ions, the ratio of the observed intensities of the masses of the plurality of ions, or a value obtained by calculation using at least one of them.

[0012] The exploration device from the seventh perspective is the exploration device from any one of the first to sixth perspectives, and the third parameter includes at least one of the film thickness, the processing shape, the processing speed, the sputter resistance, and the deposited film composition.

[0013] The search device of the eighth aspect is any one of the search devices from the first aspect to the seventh aspect. The control unit acquires each value of the second parameter and each value of the third parameter for each value of the predicted first parameter when each value of the predicted first parameter is set in the plasma processing device. The control unit updates the learning data by adding data associating each value of the predicted first parameter, each value of the acquired second parameter, and each value of the acquired third parameter to the learning data. The control unit updates the learning model using the updated learning data.

[0014] With such a configuration, the search device of the eighth aspect can further improve the prediction accuracy of the learning model.

[0015] The plasma processing apparatus according to the ninth aspect processes a processing target into a target object using a process gas. The plasma processing apparatus includes a processing chamber, a plasma generation device, and a search device. The processing target is placed in the processing chamber. The plasma generation device generates a plasma-processed process gas in the processing chamber according to the setting of each value of one or more first parameters, and performs plasma processing on the processing target. The search device searches for each value of the first parameters to be set in the plasma generation device. The search device has a control unit and a storage unit. The control unit selects a plurality of first combinations composed of values that each first parameter can take. When each combination of the first combinations is set in the plasma generation device, the control unit acquires, for each combination of the first combinations, each value of one or more second parameters indicating the state of the plasma-processed process gas and each value of one or more third parameters indicating the evaluation of the processed processing target after processing. The control unit stores learning data associating the combination of the first combinations, each value of the acquired second parameters, and each value of the acquired third parameters in the storage unit. The control unit generates a learning model using the learning data. The learning model uses the first parameters and the second parameters as explanatory variables and the third parameter as an objective variable. The control unit predicts each value of the first parameters such that each value of the third parameter becomes a predetermined value using the learning model.

[0016] In the plasma processing apparatus according to the ninth aspect, the learning model uses, as explanatory variables, one or more second parameters indicating the state of the plasma-processed process gas in addition to one or more first parameters to be set in the plasma generation device. As a result, the plasma processing apparatus can further improve the prediction accuracy of the learning model.

[0017] The search method of the tenth aspect searches for the respective values of one or more first parameters, which are set in a plasma processing apparatus that processes a processing target into a target object using a process gas. The search method includes a selection step, an acquisition step, a generation step, and a prediction step. The selection step selects a plurality of first combinations each composed of values that each first parameter can take. The acquisition step acquires, for each combination of the first combinations, the respective values of one or more second parameters indicating the state of the plasma-processed process gas and the respective values of one or more third parameters indicating the evaluation of the processed processing target, when each combination of the first combinations is set in the plasma processing apparatus. The generation step stores, in a storage unit, learning data in which the combinations of the first combinations, the respective values of the acquired second parameters, and the respective values of the acquired third parameters are associated. The generation step generates a learning model using the learning data. The learning model has the first parameter and the second parameter as explanatory variables and the third parameter as an objective variable. The prediction step predicts the respective values of the first parameter such that the respective values of the third parameter become a predetermined value, using the learning model.

[0018] In the search method of the tenth aspect, the learning model has, as explanatory variables, one or more second parameters indicating the state of the plasma-processed process gas, in addition to one or more first parameters set in the plasma processing apparatus. As a result, the search method can further improve the prediction accuracy of the learning model.

Brief Description of the Drawings

[0019]

Figure 1

Figure 2A

Figure 2B

Figure 2C

Figure 3

Embodiments for Carrying Out the Invention

[0020] (1) Overall Configuration The plasma processing apparatus 1 processes a processing target into a target object using a process gas. The plasma processing apparatus 1 is, for example, a film forming apparatus, a pattern processing apparatus, an ion implantation apparatus, etc. The film forming apparatus is, for example, CVD (Chemical Vapor Deposition), PVD (Physical Vapor Deposition), etc. The pattern processing apparatus is, for example, a dry etching apparatus, etc. The ion implantation apparatus is, for example, a plasma doping apparatus, etc.

[0021] FIG. 1 is a functional block diagram of the plasma processing apparatus 1. As shown in FIG. 1, the plasma processing apparatus 1 mainly includes a processing chamber 70, a plasma generation apparatus 2, and a search apparatus 3. The plasma generation apparatus 2 and the search apparatus 3 are communicably connected via a communication line 90.

[0022] (2) Detailed Configuration (2-1) Processing Chamber The processing target is placed in the processing chamber 70.

[0023] In the processing chamber 70, various measuring devices for measuring the state of the plasma-processed process gas and the state of the processed processing target after processing are installed. The measuring devices are, for example, a mass spectrometer, an emission spectroscope, a film thickness meter, an SEM (Scanning Electron Microscope), an XPS (X-ray Photoelectron Spectroscopy), a temperature sensor, a pressure sensor, and a flow sensor, etc.

[0024] (2-2) Plasma Generation Apparatus The plasma generation device 2 generates a plasma-processed gas in the processing chamber 70 according to the setting of each value of one or more first parameters, and performs plasma processing on the object to be processed. Hereinafter, one combination of each value of one or more first parameters may be described as a recipe. As shown in FIG. 1, the plasma generation device 2 mainly includes a storage unit 21, a communication unit 22, and a control unit 29. The control unit 29, the storage unit 21, and the communication unit 22 are communicably connected via internal wiring.

[0025] (2-2-1) Storage Unit The storage unit 21 is a storage device such as a RAM, a ROM, and an HDD (hard disk drive). The storage unit 21 stores programs executed by the control unit 29, data necessary for the execution of the programs, and the like.

[0026] (2-2-2) Communication Unit The communication unit 22 is a communication interface device for communicating with the exploration device 3 via the communication line 90.

[0027] (2-2-3) Control Unit The control unit 29 is a processor such as a CPU or a GPU. The control unit 29 reads and executes the programs stored in the storage unit 21 to realize various functions of the plasma generation device 2. The control unit 19 can write the calculation results to the storage unit 21 or read the information stored in the storage unit 21 according to the programs.

[0028] As shown in FIG. 1, the control unit 29 mainly includes an acquisition unit 291 and a plasma processing unit 292 as functional blocks.

[0029] (2-2-3-1) Acquisition Unit The acquisition unit 291 acquires each value of one or more first parameters from the exploration device 3. The acquisition unit 291 sets each value of the first parameter acquired from the exploration device 3 to the plasma processing unit 292.

[0030] The first parameter includes at least one of plasma generation power, bias power, pressure of the process gas, flow rate of the process gas, and type of the process gas. The plasma generation power is, for example, ICP (Inductively Coupled Plasma) power, CCP (Capacitively Coupled Plasma) power, ECR (Electron Cyclotron Resonance) power, or the like. The value of the type of the process gas is, for example, a fluorine-based gas such as CF4, CHF3, or a mixed gas of a fluorine-based gas such as CF4 + H2 and hydrogen. The first parameter may further include the ratio of hydrogen contained in the process gas.

[0031] The acquisition unit 291 acquires, using various measuring devices installed in the processing chamber 70, the respective values of one or more second parameters indicating the state of the plasma-processed process gas and the respective values of one or more third parameters indicating the evaluation of the object to be processed after processing. The acquisition unit 291 transmits the acquired respective values of the second parameters and the respective values of the third parameters to the search device 3.

[0032] The second parameter includes a parameter directly representing at least one of electron temperature, electron density, radical species, radical density, ion species, ion density, and ion energy. The acquisition unit 291 acquires electron temperature, electron density, radical species, radical density, ion species, ion density, and ion energy using, for example, a mass spectrometer and an emission spectroscope.

[0033] The second parameter may include a parameter that indirectly represents at least one of electron temperature, electron density, radical species, radical density, ion species, ion density, and ion energy. For example, the parameter that indirectly represents radical species and the parameter that indirectly represents radical density include at least one of the emission wavelength of the plasma, the emission intensity of the emission wavelength of the plasma, the ratio of the emission intensities of the plurality of emission wavelengths of the plasma, the mass of the radicals in the plasma, the observed intensity of the mass of the radicals in the plasma, the ratio of the observed intensities of the masses of the plurality of radicals in the plasma, or a value obtained by calculation using at least one of them. For example, the parameter that indirectly represents ion species and the parameter that indirectly represents ion density include at least one of the emission wavelength of the plasma, the emission intensity of the emission wavelength of the plasma, the ratio of the emission intensities of the plurality of emission wavelengths of the plasma, the mass of the ions, the observed intensity of the mass of the ions, the ratio of the observed intensities of the masses of the plurality of ions, or a value obtained by calculation using at least one of them. The acquisition unit 291 acquires the emission wavelength of the plasma and the emission intensity of the emission wavelength of the plasma using, for example, an emission spectroscope or the like. The acquisition unit 291 acquires the mass of the radicals in the plasma, the observed intensity of the mass of the radicals in the plasma, the mass of the ions, and the observed intensity of the mass of the ions using, for example, a mass analyzer or the like.

[0034] The third parameter includes at least one of film thickness, processed shape, processing speed, sputter resistance, and deposited film composition. The acquisition unit 291 acquires the value of the film thickness using, for example, a film thickness meter, SEM, or the like. The processing speed is, for example, an etching rate, a film formation rate, or the like. The acquisition unit 291 acquires the value of the processing speed using, for example, a flow rate sensor or the like. The processed shape is, for example, processing depth, selectivity, maximum diameter, bottom diameter, and blockage. The acquisition unit 291 acquires the value of the processed shape using, for example, SEM or the like. The acquisition unit 291 acquires the values of sputter resistance and deposited film composition using, for example, XPS or the like.

[0035] (2-2-3-2) Plasma processing unit The plasma processing unit 292 generates a plasma-processed process gas in the processing chamber 70 according to the recipe set by the acquisition unit 291, and performs plasma processing on the object to be processed.

[0036] (2-3) Exploration device The exploration device 3 explores the recipe to be set in the plasma generation device 2. As shown in FIG. 1, the exploration device 3 mainly includes a storage unit 31, a communication unit 32, an input unit 33, a display unit 34, and a control unit 39. The control unit 39, the storage unit 31, the communication unit 32, the input unit 33, and the display unit 34 are communicably connected via internal wiring.

[0037] (2-3-1) Storage unit The storage unit 31 is a storage device such as a RAM, a ROM, and an HDD (hard disk drive). The storage unit 31 stores programs executed by the control unit 39 and data necessary for the execution of the programs.

[0038] The storage unit 31 stores learning data 81 and a learning model 82, which will be described later. Further, the storage unit 31 stores, for example, commercially available software for realizing the experimental design method and machine learning.

[0039] (2-3-2) Communication unit The communication unit 32 is a communication interface device for communicating with the plasma generation device 2 via the communication line 90.

[0040] (2-3-3) Input unit The input unit 33 is, for example, a keyboard and a mouse. The user can input various commands and various information into the exploration device 3 using the input unit 33.

[0041] The user inputs the values that each first parameter can take (the domain of each first parameter) using the input unit 33.

[0042] (2-3-4) Display unit The display unit 34 is, for example, a monitor. On the display unit 34, a screen for inputting various types of information and various types of information stored in the storage unit 31 and the like are displayed.

[0043] On the display unit 34, for example, the finally selected recipe is displayed.

[0044] (2-3-5) Control unit The control unit 39 is a processor such as a CPU or a GPU. The control unit 39 reads and executes the programs stored in the storage unit 31 to realize various functions of the search device 3. The control unit 39 can write the calculation results to the storage unit 31 or read the information stored in the storage unit 31 according to the programs.

[0045] As shown in FIG. 1, the control unit 39 mainly includes, as functional blocks, an acquisition unit 391, a selection unit 392, a generation unit 393, and a prediction unit 394.

[0046] (2-3-5-1) Acquisition unit The acquisition unit 391 acquires the values that each first parameter can take from the user via the input unit 33. The acquisition unit 291 transmits the values that each first parameter can take to the selection unit 392.

[0047] When the acquisition unit 391 receives the first recipe group from the selection unit 392, it transmits the first recipe group to the plasma generation device 2. When each recipe of the first recipe group is set in the plasma generation device 2 by the acquisition unit 391, for each recipe of the first recipe group, it acquires the respective values of one or more second parameters and the respective values of one or more third parameters. The acquisition unit 391 transmits the first recipe group received from the selection unit 392, the respective values of the second parameters, and the respective values of the third parameters acquired from the plasma generation device 2 to the selection unit 392.

[0048] When the acquisition unit 391 receives the first predicted recipe or the second predicted recipe from the prediction unit 394, it transmits the first predicted recipe or the second predicted recipe to the plasma generation device 2. When each recipe of the first predicted recipe or the second predicted recipe is set in the plasma generation device 2, the acquisition unit 391 acquires each value of the second parameter and each value of the third parameter for each recipe of the first predicted recipe or the second predicted recipe. The acquisition unit 391 transmits the first predicted recipe or the second predicted recipe received from the prediction unit 394, each value of the second parameter, and each value of the third parameter acquired from the plasma generation device 2 to the selection unit 392.

[0049] (2-3-5-2) Selection Unit When the selection unit 392 receives the possible values that each first parameter can take from the acquisition unit 391, it selects a first recipe group (a plurality of first combinations) composed of the possible values that each first parameter can take.

[0050] Specifically, first, the selection unit 392 defines a second recipe group (a plurality of second combinations) composed of the possible values that each first parameter can take. For example, the second recipe group is a set of all combinations composed of the possible values that each first parameter can take.

[0051] Next, the selection unit 392 selects a first recipe group with a smaller number of recipes than the second recipe group from the second recipe group by means of the experimental design method. The experimental design method is, for example, the D-optimal design, the orthogonal array, the central composite design, etc. The selection unit 392 executes the experimental design method by using, for example, commercially available software for realizing the experimental design method.

[0052] The selection unit 392 transmits the selected first recipe group to the acquisition unit 391.

[0053] When the selection unit 392 receives from the acquisition unit 391 the first recipe group, the respective values of the second parameter, and the respective values of the third parameter, it determines whether there is a recipe in the first recipe group such that the respective values of the third parameter become the first predetermined value. The first predetermined value is the respective value of the third parameter that the target object has. If it exists, the selection unit 392 selects the recipe. In other words, the recipe is a recipe for processing the object to be processed into the target object. If it does not exist, the selection unit 392 transmits to the generation unit 393 the first recipe group, the respective values of the second parameter, and the respective values of the third parameter received from the acquisition unit 391.

[0054] When the selection unit 392 receives from the acquisition unit 391 the first prediction recipe or the second prediction recipe, the respective values of the second parameter, and the respective values of the third parameter, it determines whether there is a recipe in the first prediction recipe or the second prediction recipe such that the respective values of the third parameter become the first predetermined value. If it exists, the selection unit 392 selects the recipe. In other words, the recipe is a recipe for processing the object to be processed into the target object. If it does not exist, the selection unit 392 transmits to the generation unit 393 the first prediction recipe or the second prediction recipe, the respective values of the second parameter, and the respective values of the third parameter received from the acquisition unit 391.

[0055] (2-3-5-3) Generation unit When the generation unit 393 receives from the selection unit 392 the first recipe group, the respective values of the second parameter, and the respective values of the third parameter, it stores in the storage unit 31 the learning data 81 in which the recipes of the first recipe group, the respective values of the second parameter, and the respective values of the third parameter are associated. The generation unit 393 generates a learning model 82 using the learning data 81. The learning model 82 has the first parameter and the second parameter as explanatory variables and the third parameter as an objective variable. The learning model 82 is, for example, a Gaussian process regression, a linear regression, a regression tree, or the like. The generation unit 393 generates the learning model 82 using, for example, commercially available software for realizing machine learning.

[0056] When the generation unit 393 receives the first prediction recipe or the second prediction recipe from the selection unit 392, the generation unit 393 updates the learning data 81 by adding data in which each recipe of the first prediction recipe or the second prediction recipe is associated with each value of the second parameter and each value of the third parameter to the learning data 81. The generation unit 393 updates the learning model 82 using the updated learning data 81.

[0057] (2-3-5-4) Prediction Unit When the generation unit 393 generates or updates the learning model 82, the prediction unit 394 uses the learning model 82 to predict a recipe in which each value of the third parameter becomes a first predetermined value.

[0058] Specifically, the prediction unit 394 defines a third recipe group having a larger number of recipes than the second recipe group, and inputs each recipe of the third recipe group into the learning model 82. For example, the third recipe group is a set of all combinations composed of values that each first parameter can take after increasing the values that each first parameter can take. At this time, each value of the second parameter input to the learning model 82 is predicted from each recipe of the third recipe group, for example. For example, the generation unit 393 generates a learning model having the first parameter as an explanatory variable and the second parameter as an objective variable using the learning data in which the recipes of the first recipe group are associated with the respective values of the second parameter. The prediction unit 394 predicts each value of the second parameter input to the learning model 82 by inputting each recipe of the third recipe group into the learning model.

[0059] Each value of the second parameter input to the learning model 82 may be, for example, a pre-defined value. For example, when the second parameter is the ion energy, five levels of values, (100, 150, 200, 250, 300) (unit: eV), are pre-defined as the ion energy. The prediction unit 394 inputs five values (five levels) of the ion energy to the learning model 82 for each recipe in the third recipe group.

[0060] The prediction unit 394 determines whether there is a recipe such that each value of the third parameter output from the learning model 82 becomes the first predetermined value. If it exists, the prediction unit 394 transmits to the acquisition unit 391 a recipe (first prediction recipe) such that each value of the third parameter becomes the first predetermined value. If it does not exist, the prediction unit 394 extracts one or more recipes (second prediction recipes) from the set obtained by removing the recipes included in the current learning data 81 from the third recipe group, and transmits the second prediction recipes to the acquisition unit 391. For example, when the learning model 82 is Gaussian process regression, the prediction unit 394 extracts the second prediction recipes using Bayesian optimization.

[0061] (3) Processing An example of the processing of the plasma processing apparatus 1 will be described using the flowcharts of FIGS. 2A to 2C.

[0062] As shown in step S1, the user inputs the definition domain of each first parameter to the search device 3.

[0063] After finishing step S1, as shown in step S2, the search device 3 defines the second recipe group, and selects the first recipe group from the second recipe group by the experimental design method.

[0064] After finishing step S2, as shown in step S3, the search device 3 transmits the first recipe group to the plasma generation device 2.

[0065] After finishing step S3, as shown in step S4, the plasma generation device 2 sets the first recipe group in the plasma processing unit 292 and performs plasma processing. The plasma generation device 2 transmits the respective values of the second parameter and the respective values of the third parameter obtained by the plasma processing to the search device 3.

[0066] After finishing step S4, as shown in step S5, the search device 3 determines whether there is a recipe in the first recipe group such that the respective values of the third parameter become the first predetermined value. If it exists, the process proceeds to step S14. If it does not exist, the process proceeds to step S6.

[0067] When proceeding from step S5 to step S6, the search device 3 stores learning data 81 associating the recipes in the first recipe group, the respective values of the second parameter, and the respective values of the third parameter in the storage unit 31. The search device 3 generates a learning model 82 using the learning data 81.

[0068] After finishing step S6, as shown in step S7, the search device 3 defines the third recipe group and inputs each recipe in the third recipe group into the learning model 82.

[0069] After finishing step S7, as shown in step S8, the search device 3 determines whether there is a recipe such that the respective values of the third parameter output from the learning model 82 become the first predetermined value. If it exists, the process proceeds to step S9. If it does not exist, the process proceeds to step S10.

[0070] When proceeding from step S8 to step S9, the search device 3 transmits the first predicted recipe to the plasma generation device 2.

[0071] When proceeding from step S8 to step S10, the search device 3 transmits the second predicted recipe to the plasma generation device 2.

[0072] When step S9 or step S10 is completed, as shown in step S11, the plasma generation device 2 sets the first prediction recipe or the second prediction recipe in the plasma processing unit 292 and performs plasma processing. The plasma generation device 2 transmits the respective values of the second parameters and the respective values of the third parameters obtained by the plasma processing to the search device 3.

[0073] When step S11 is completed, as shown in step S12, the search device 3 determines whether there is a recipe of the first prediction recipe or the second prediction recipe such that the respective values of the third parameters become the first predetermined value. If it exists, the process proceeds to step S14. If it does not exist, the process proceeds to step S13.

[0074] When proceeding from step S12 to step S13, the search device 3 updates the learning data 81 by adding data associating each recipe of the first prediction recipe or the second prediction recipe, each value of the second parameter, and each value of the third parameter to the learning data 81. The generation unit 393 updates the learning model 82 using the updated learning data 81.

[0075] When step S13 is completed, the process returns to step S7, and the search device 3 inputs each recipe of the third recipe group into the learning model 82.

[0076] When proceeding from step S5 or step S12 to step S14, the search device 3 selects a recipe such that the value of the third parameter becomes the first predetermined value. In other words, the search device 3 selects a recipe for processing the object to be processed into the target object.

[0077] (4) Verification (4-1) Premise In this verification, until a recipe for processing the object to be processed into the target object is selected, it is verified how many times the plasma processing apparatus 1 of the present embodiment performs plasma processing (the number of recipes set in the plasma generation device 2).

[0078] This verification shows that the plasma processing apparatus 1 is a dry etching apparatus. In other words, the plasma processing apparatus 1 processes the object to be processed into a target object by plasma processing using an etching gas (process gas).

[0079] The source power and the bottom power supplied to the processing chamber 70 are high-frequency power sources of 13.56 MHz.

[0080] The object to be processed is SiO2 (TEOS).

[0081] The first parameters are ICP power, bias power, the pressure of the etching gas, the flow rate of the etching gas, the ratio of hydrogen contained in the etching gas, and the type of the etching gas.

[0082] As the definition domain of the first parameters, the ICP power has four levels of values: (400, 600, 800, 1000) (unit: W). The bias power has three levels of values: (100, 200, 300) (unit: W). The pressure of the etching gas has four levels of values: (10, 20, 30, 40) (unit: mTorr). The flow rate of the etching gas has three levels of values: (50, 75, 100) (unit: sccm). The ratio of hydrogen contained in the etching gas has four levels of values: (0, 0.25, 0.5, 0.75). The type of the etching gas has four levels of values: (CF4, CF4 + H2, CF4 + H2, CF4 + H2), corresponding to the respective values of the ratio of hydrogen contained in the etching gas.

[0083] The second parameter is the observed intensity of H2F+ ions (mass m / z = 21).

[0084] The third parameter is the etching rate.

[0085] The second recipe group is a set of all combinations composed of the values that each first parameter can take. Since the ICP power, the bias power, the pressure of the etching gas, the flow rate of the etching gas, and the proportion of hydrogen contained in the etching gas have values of 4 levels, 3 levels, 4 levels, 3 levels, and 4 levels respectively, the second recipe group is composed of 576 (= 4×3×4×3×4) recipes. Note that since the value of the proportion of hydrogen contained in the etching gas and the value of the type of the etching gas correspond one-to-one, the type of the etching gas is omitted.

[0086] The third recipe group is all combinations composed of the values that each first parameter can take when the ICP power has a 7-level value of (400, 500, 600, 700, 800, 900, 1000) (unit: W), the bias power has a 5-level value of (100, 150, 200, 250, 300) (unit: W), the pressure of the etching gas has a 7-level value of (10, 15, 20, 25, 30, 35, 40) (unit: mTorr), the flow rate of the etching gas has a 6-level value of (50, 60, 70, 80, 90, 100) (unit: sccm), and the proportion of hydrogen contained in the etching gas has a 4-level value of (0, 0.25, 0.5, 0.75). The third recipe group is composed of 5880 (= 7×5×7×6×4) recipes.

[0087] The experimental design method is a D-optimal design.

[0088] The learning model 82 is a Gaussian process regression.

[0089] Since the learning model 82 is a Gaussian process regression, the prediction unit 394 extracts the second predicted recipe in order from the recipes in the set obtained by removing the recipes included in the current-stage learning data 81 from the third recipe group, among the recipes where the surface or curve showing the expected value is close to the region of the first predetermined value.

[0090] (4-2) Results As a result, out of a total of 576 recipes, only 45 recipes were set in the plasma generation device 2, and a recipe for processing the object to be processed into the target object was selected. Among the 45 recipes, 2 recipes were the ones selected for processing the object to be processed into the target object, and 43 recipes were the ones used for the learning data 81.

[0091] Also, the accuracy of the final learning model 82 (first learning model 821) and the second learning model 822 generated using the learning data obtained by removing the second parameter from the final learning data 81 was compared. FIG. 3 is a diagram showing the comparison of the accuracy of the learning models. The right diagram in FIG. 3 shows the accuracy of the first learning model 821, and the left diagram in FIG. 3 shows the accuracy of the second learning model 822. The plotted points correspond to the 43 recipes included in the final learning data 81. The horizontal axis is the etching rate obtained by setting the recipe in the plasma generation device 2. The vertical axis is the etching rate predicted by inputting the recipe into the learning model (in the first learning model 821 in the right diagram, in addition to the recipe, the value of the second parameter is also input into the learning model). Therefore, it is desirable for the plotted points to be arranged on the diagonal line.

[0092] As shown in FIG. 3, it can be seen that the plotted points by the first learning model 821 are arranged more on the diagonal line than the plotted points by the second learning model 822. The coefficient of determination was 0.9347 for the first learning model 821 and 0.8885 for the second learning model 822.

[0093] (5) Features (5-1) Conventionally, there is a technique of generating a learning model with the parameters set in the plasma processing device as explanatory variables and the parameters indicating the evaluation of the object to be processed after processing as target variables, and using the learning model to search for the values of the parameters to be set in the plasma processing device for processing the object to be processed into the target object.

[0094] In order to reduce the number of times of searching for the values of the parameters to be set in the plasma processing device, it is desirable to further improve the prediction accuracy of the learning model.

[0095] The search device 3 of this embodiment searches for the value of each of one or more first parameters for setting the plasma processing device 1 that processes a processing target into a target object using a process gas. The search device 3 includes a control unit 39 and a storage unit 31. The control unit 39 selects a plurality of first combinations composed of values that each first parameter can take. When each combination of the first combination is set in the plasma processing device 1, the control unit 39 obtains, for each combination of the first combination, the value of each of one or more second parameters indicating the state of the plasma-processed process gas and the value of each of one or more third parameters indicating the evaluation of the processed processing target after processing. The control unit 39 stores learning data 81 in which the combination of the first combination, the value of each of the obtained second parameters, and the value of each of the obtained third parameters are associated with each other in the storage unit 31. The control unit 39 generates a learning model 82 using the learning data 81. The learning model 82 uses the first parameter and the second parameter as explanatory variables and the third parameter as a target variable. The control unit 39 predicts the value of each first parameter such that the value of each third parameter becomes a predetermined value using the learning model 82.

[0096] In the search device 3 of this embodiment, the learning model 82 uses, as explanatory variables, one or more second parameters indicating the state of the plasma-processed process gas in addition to one or more first parameters set in the plasma processing device 1. As a result, the search device 3 can further improve the prediction accuracy of the learning model 82.

[0097] (5-2) In the search device 3 of this embodiment, the control unit 39 selects a first combination having a smaller number of combinations than the second combination from among a plurality of second combinations composed of values that each first parameter can take by the experimental design method.

[0098] As a result, the search device 3 can efficiently search for the value of the parameter set in the plasma processing device.

[0099] (5-3) In the search device 3 of the present embodiment, the first parameter includes at least one of the power for plasma generation, the bias power, the pressure of the process gas, the flow rate of the process gas, and the type of the process gas.

[0100] (5-4) In the search device 3 of the present embodiment, the second parameter includes a parameter that directly or indirectly represents at least one of the electron temperature, the electron density, the radical species, the radical density, the ion species, the ion density, and the ion energy.

[0101] (5-5) In the search device 3 of the present embodiment, the parameter that indirectly represents the radical species and the parameter that indirectly represents the radical density include at least one of the emission wavelength of the plasma, the emission intensity of the emission wavelength of the plasma, the ratio of the emission intensities of the plurality of emission wavelengths of the plasma, the mass of the radical in the plasma, the observed intensity of the mass of the radical in the plasma, the ratio of the observed intensities of the masses of the plurality of radicals in the plasma, or a value obtained by calculation using at least one of them.

[0102] (5-6) In the search device 3 of the present embodiment, the parameter that indirectly represents the ion species and the parameter that indirectly represents the ion density include at least one of the emission wavelength of the plasma, the emission intensity of the emission wavelength of the plasma, the ratio of the emission intensities of the plurality of emission wavelengths of the plasma, the mass of the ion, the observed intensity of the mass of the ion, the ratio of the observed intensities of the masses of the plurality of ions, or a value obtained by calculation using at least one of them.

[0103] (5-7) In the search device 3 of the present embodiment, the third parameter includes at least one of the film thickness, the processing shape, the processing speed, the sputter resistance, and the deposited film composition.

[0104] (5-8) In the search device 3 of the present embodiment, when each value of the predicted first parameter is set in the plasma processing device 1, the control unit 39 acquires each value of the second parameter and each value of the third parameter for the value of the predicted first parameter. The control unit 39 updates the learning data 81 by adding data associating each value of the predicted first parameter, each value of the acquired second parameter, and each value of the acquired third parameter to the learning data 81. The control unit 39 updates the learning model 82 using the updated learning data 81.

[0105] As a result, the search device 3 can further improve the prediction accuracy of the learning model 82.

[0106] (5-9) The plasma processing apparatus 1 of the present embodiment processes a processing target into a target object using a process gas. The plasma processing apparatus 1 includes a processing chamber 70, a plasma generation device 2, and a search device 3. The processing target is placed in the processing chamber 70. The plasma generation device 2 generates a process gas converted into plasma in the processing chamber 70 according to the setting of each value of one or more first parameters, and performs plasma processing on the processing target. The search device 3 searches for each value of the first parameters to be set in the plasma generation device 2. The search device 3 includes a control unit 39 and a storage unit 31. The control unit 39 selects a plurality of first combinations each composed of values that each first parameter can take. When each combination of the first combinations is set in the plasma generation device 2, the control unit 39 acquires, for each combination of the first combinations, each value of one or more second parameters indicating the state of the plasma-converted process gas and each value of one or more third parameters indicating the evaluation of the processed processing target after processing. The control unit 39 stores learning data 81 associating the combination of the first combinations, each value of the acquired second parameters, and each value of the acquired third parameters in the storage unit 31. The control unit 39 generates a learning model 82 using the learning data 81. The learning model 82 has the first parameters and the second parameters as explanatory variables and the third parameter as an objective variable. The control unit 39 predicts each value of the first parameters such that each value of the third parameter becomes a predetermined value using the learning model 82.

[0107] In the plasma processing apparatus 1 of the present embodiment, the learning model 82 uses, as explanatory variables, one or more second parameters indicating the state of the plasma-converted process gas in addition to one or more first parameters to be set in the plasma generation device 2. As a result, the plasma processing apparatus 1 can further improve the prediction accuracy of the learning model 82.

[0108] (5-10) The search method of this embodiment searches for the respective values of one or more first parameters set in the plasma processing apparatus 1 that processes a target object into a target using a process gas. The search method includes a selection step S2, an acquisition step S4, a generation step S6, and prediction steps S7 and S8. The selection step S2 selects a plurality of first combinations each composed of values that each first parameter can take. In the acquisition step S4, when each combination of the first combination is set in the plasma processing apparatus 1, for each combination of the first combination, the respective values of one or more second parameters indicating the state of the plasma-processed process gas and the respective values of one or more third parameters indicating the evaluation of the processed target object after processing are acquired. The generation step S6 stores learning data 81 in which the combination of the first combination, the respective values of the acquired second parameters, and the respective values of the acquired third parameters are associated in the storage unit 31. The generation step S6 generates a learning model 82 using the learning data 81. The learning model 82 uses the first parameter and the second parameter as explanatory variables and the third parameter as an objective variable. The prediction steps S7 and S8 predict the respective values of the first parameter such that the respective values of the third parameter become a predetermined value using the learning model 82.

[0109] In the search method of this embodiment, the learning model 82 uses, as explanatory variables, one or more second parameters indicating the state of the plasma-processed process gas in addition to one or more first parameters set in the plasma processing apparatus 1. As a result, the search method can further improve the prediction accuracy of the learning model 82.

[0110] (6) Variation (6-1) Variation 1A In this embodiment, the plasma processing apparatus 1 had the search apparatus 3. However, the search apparatus 3 may be separated from the plasma processing apparatus 1. In this case, the communication line 90 is, for example, the Internet.

[0111] (6-2) As described above, although embodiments of the present disclosure have been explained, it will be understood that various changes in form and details are possible without departing from the spirit and scope of the present disclosure described in the claims.

Explanation of Signs

[0112] 1 Plasma processing apparatus 2 Plasma generation apparatus 3 Search apparatus 31 Storage unit 39 Control unit 70 Processing chamber 81 Learning data 82 Learning model S2 Selection step S4 Acquisition step S6 Generation step S7, S8 Prediction steps

Prior Art Documents

Patent Documents

[0113]

Patent Document 1

Claims

1. A search device (3) for searching for the value of each of one or more first parameters, which is set to a plasma processing device (1) that processes an object to be processed into a target object using a process gas, comprising a control unit (39) and a storage unit (31), wherein the control unit selects a plurality of first combinations each composed of values that each of the first parameters can take, and when each combination of the first combinations is set in the plasma processing device, for each combination of the first combinations, the value of each of one or more second parameters indicating the state of the plasma-processed process gas and the value of each of one or more third parameters indicating the evaluation of the processed object after processing are acquired, stores learning data (81) associating the combination of the first combinations, the value of each of the acquired second parameters, and the value of each of the acquired third parameters in the storage unit, and generates a learning model (82) using the learning data with the first parameter and the second parameter as explanatory variables and the third parameter as an objective variable, and predicts the value of each of the first parameters such that the value of each of the third parameters becomes a predetermined value using the learning model, search device (3).

2. The control unit selects, by means of an experimental design method, the first combination having a smaller number of combinations than the second combination from among a plurality of second combinations each composed of values that each of the first parameters can take, The search device (3) according to claim 1.

3. The first parameter includes at least one of plasma generation power, bias power, the pressure of the process gas, the flow rate of the process gas, and the type of the process gas, The search device (3) according to claim 1 or 2.

4. The second parameter includes a parameter that directly or indirectly represents at least one of electron temperature, electron density, radical species, radical density, ion species, ion density, and ion energy, The search device (3) according to claim 1 or 2.

5. The parameter that indirectly represents the radical species and the parameter that indirectly represents the radical density are the emission wavelength of the plasma and the emission intensity of the emission wavelength of the plasma, and the ratio of the emission intensities of the plurality of emission wavelengths of the plasma, ​ The mass of radicals in the plasma, and the observed intensity of the mass of radicals in the plasma, and the ratio of the observed intensities of the masses of a plurality of radicals in the plasma, at least one of the above, or a value obtained by performing an operation using at least one of the above, The search device according to claim 4.

6. The parameter indirectly representing the ion species and the parameter indirectly representing the ion density are the emission wavelength of the plasma and the emission intensity of the emission wavelength of the plasma, and the ratio of the emission intensities of a plurality of emission wavelengths of the plasma, the mass of the ion and the observed intensity of the mass of the ion, and the ratio of the observed intensities of the masses of a plurality of ions, at least one of the above, or a value obtained by performing an operation using at least one of the above, The search device according to claim 4.

7. The third parameter includes at least one of film thickness, processed shape, processing speed, sputter resistance, and deposited film composition. The search device (3) according to claim 1 or 2.

8. The control unit when each predicted value of the first parameter is set in the plasma processing apparatus, for each predicted value of the first parameter, obtains each value of the second parameter and each value of the third parameter, updates the learning data by adding data associating each predicted value of the first parameter, each obtained value of the second parameter, and each obtained value of the third parameter to the learning data, and updates the learning model using the updated learning data. The search device (3) according to claim 1 or 2.

9. A plasma processing apparatus (1) for processing a processing target into a target object using a process gas, a processing chamber (70) on which the processing target is placed, a plasma generation device (2) that generates the process gas converted into plasma in the processing chamber according to the setting of each value of one or more first parameters and performs plasma processing on the processing target, a search device (3) that searches for each value of the first parameter to be set in the plasma generation device, comprising the search device has a control unit (39) and a storage unit (31), the control unit selects a plurality of first combinations each composed of values that each of the first parameters can take, By setting each combination of the first combination in the plasma generation device, for each combination of the first combination, each value of one or more second parameters indicating the state of the processed gas that has been turned into plasma and each value of one or more third parameters indicating the evaluation of the object to be processed after processing are acquired. Learning data (81) associating the combination of the first combination, each value of the acquired second parameters, and each value of the acquired third parameters is stored in the storage unit. Using the learning data, a learning model (82) is generated with the first parameter and the second parameter as explanatory variables and the third parameter as the objective variable. Using the learning model, each value of the first parameter is predicted such that each value of the third parameter becomes a predetermined value. Plasma processing apparatus (1).

10. A search method for searching for each value of one or more first parameters to be set in a plasma processing apparatus (1) that processes an object to be processed into a target object using a process gas, A selection step (S2) of selecting a plurality of first combinations composed of values that each of the first parameters can take; By setting each combination of the first combination in the plasma processing apparatus, for each combination of the first combination, each value of one or more second parameters indicating the state of the processed gas that has been turned into plasma and each value of one or more third parameters indicating the evaluation of the object to be processed after processing are acquired. An acquisition step (S4); Learning data associating the combination of the first combination, each value of the acquired second parameters, and each value of the acquired third parameters is stored in the storage unit. Using the learning data, a learning model is generated with the first parameter and the second parameter as explanatory variables and the third parameter as the objective variable. A generation step (S6); Using the learning model, each value of the first parameter is predicted such that each value of the third parameter becomes a predetermined value. A prediction step (S7, S8); Comprising: Search method.

Images