Learning device, information processing device, substrate processing system, learning method, recipe determination method and learning program

Abstract

To provide a learning device capable of generating a prediction model for searching an optimal processing condition of a substrate even in the case that there are deficiencies in the physical property value of a subliming agent or solvent, an information processing device, a substrate processing system, a learning method, a recipe determination method and a learning program.SOLUTION: A data acquisition part acquires a processing condition in processing for sublimating and drying a substrate by using process liquid containing a subliming agent and solvent and data indicating a relation between at least one molecular descriptor of the subliming agent and the solvent, and a processing result of the substrate. A data complement part complements deficiencies in the physical property value of at least one of the subliming agent and the solvent in the data acquired by the data acquisition part. A result model generation part 412 generates a result model for predicting a processing result of the substrate by performing machine learning the data obtained by complementing the physical property value by the data complement part.SELECTED DRAWING: Figure 6

Description

Technical Field

[0001] The present invention relates to a learning device, an information processing device, a substrate processing system, a learning method, a recipe determination method, and a learning program.

Background Art

[0002] In recent years, due to the miniaturization of patterns formed on a substrate, the pattern strength tends to become weak. In this case, when drying the liquid applied to the substrate, the pattern is likely to collapse due to the surface tension acting between the surface of the pattern formed on the substrate and the liquid. Therefore, in order to dry well a substrate having a fine pattern formed on its surface, sublimation drying may be performed on the substrate. Sublimation drying is a technique for drying a liquid by converting the liquid applied to the surface of the substrate into a solid and then causing a phase change from the solid to a gas.

[0003] For example, in the sublimation drying described in Patent Document 1, a treatment liquid in which cyclohexanone oxime is dissolved in a solvent is supplied to the surface of a substrate on which a pattern is formed, whereby a liquid film of the treatment liquid is formed on the surface of the substrate. The treatment liquid is adjusted to a concentration suitable for treatment conditions such as the type of the substrate or the rotation speed. Next, the liquid film of the treatment liquid is solidified to form a solidified film of cyclohexanone oxime. Thereafter, the solidified film is sublimated to remove the solidified film from the surface of the substrate.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] In sublimation drying, a treatment liquid in which a sublimating agent and a solvent are appropriately combined according to the substrate to be treated is used. Here, when the sublimating agent or the solvent changes, it is necessary to change the treatment conditions of the substrate. However, it requires a large cost to confirm in advance the optimal treatment conditions for each combination of the sublimating agent and the solvent through experiments or the like.

[0006] Therefore, the present inventor considered exploring the optimal treatment conditions for any combination of a sublimating agent and a solvent by using a prediction model generated by machine learning. However, according to the findings of the present inventor, it was found that the physical property values of the sublimating agent or the solvent used in the field of sublimation drying are often not investigated. Therefore, it is difficult to explore the optimal treatment conditions using a prediction model.

[0007] An object of the present invention is to provide a learning device, an information processing device, a substrate processing system, a learning method, a recipe determination method, and a learning program capable of generating a prediction model for exploring the optimal treatment conditions of a substrate even when there are deficiencies in the physical property values of the sublimating agent or the solvent.

Means for Solving the Problems

[0008] A learning device according to an aspect of the present invention includes: a data acquisition unit that acquires data indicating the relationship between the treatment conditions in a treatment for sublimation-drying a substrate using a treatment liquid containing a sublimating agent and a solvent, at least one molecular descriptor of the sublimating agent and the solvent, and the treatment result of the substrate; a data complementation unit that complements the deficiency of at least one physical property value of the sublimating agent and the solvent in the data acquired by the data acquisition unit by using a prediction algorithm that predicts physical property values from molecular descriptors; and a result model generation unit that generates a result model for predicting the treatment result of the substrate by machine-learning the data whose physical property values have been complemented by the data complementation unit.

[0009] An information processing apparatus according to another aspect of the present invention is an information processing apparatus that uses the result model generated by the above learning apparatus, and includes a model acquisition unit that acquires the result model, a sublimating agent determination unit that determines a sublimating agent used for sublimation drying of a substrate, a solvent determination unit that determines a solvent used for sublimation drying of the substrate, a condition selection unit that selects processing conditions, and using the result model acquired by the model acquisition unit, the sublimating agent determined by the sublimating agent determination unit, the solvent determined by the solvent determination unit, and a result prediction unit that predicts a processing result of the substrate from the processing conditions selected by the condition selection unit.

[0010] A substrate processing system according to still another aspect of the present invention includes the above information processing apparatus.

[0011] A learning method according to still another aspect of the present invention includes obtaining data indicating a relationship between processing conditions in a process of sublimation drying a substrate using a processing liquid containing a sublimating agent and a solvent, at least one molecular descriptor of the sublimating agent and the solvent, and a processing result of the substrate, complementing a deficiency in at least one physical property value of the sublimating agent and the solvent in the obtained data using a prediction algorithm that predicts physical property values from molecular descriptors, and generating a result model that predicts a processing result of the substrate by machine learning the data with the physical property values complemented, and is executed by a processor.

[0012] A recipe determination method according to still another aspect of the present invention is a recipe determination method that uses the result model generated by the above learning method, and includes acquiring the result model, determining a sublimating agent used for sublimation drying of a substrate, determining a solvent used for sublimation drying of the substrate, selecting processing conditions, and predicting a processing result of the substrate from the determined sublimating agent, the determined solvent, and the selected processing conditions using the acquired result model, and is executed by a processor.

[0013] A learning program according to still another aspect of the present invention causes a processor to execute: a process of acquiring data indicating a relationship between processing conditions in a process of sublimating and drying a substrate using a processing liquid containing a sublimating agent and a solvent, and at least one molecular descriptor of the sublimating agent and the solvent, and a processing result of the substrate; a process of complementing a deficiency in at least one physical property value of the sublimating agent and the solvent in the acquired data using a prediction algorithm that predicts a physical property value from a molecular descriptor; and a process of generating a result model that predicts a processing result of a substrate by machine learning the data in which the physical property value is complemented.

Advantages of the Invention

[0014] According to the present invention, even when there is a deficiency in the physical property value of the sublimating agent or the solvent, it is possible to generate a prediction model for searching for optimal processing conditions for the substrate.

Brief Description of the Drawings

[0015]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Mode for Carrying Out the Invention

[0016] 1. Substrate Processing System Hereinafter, a substrate processing system according to an embodiment of the present invention will be described with reference to the drawings. In the following description, the substrate refers to a semiconductor substrate (wafer), a substrate for a flat panel display (FPD) such as a liquid crystal display device or an organic EL (Electro Luminescence) display device, a substrate for an optical disk, a substrate for a magnetic disk, a substrate for a magneto-optical disk, a substrate for a photomask, a ceramic substrate, or a substrate for a solar cell. FIG. 1 is a diagram showing an example of the configuration of a substrate processing system according to an embodiment of the present invention. As shown in FIG. 1, the substrate processing system 500 includes a substrate processing apparatus 100, a database storage device 200, an information processing apparatus 300, and a learning apparatus 400.

[0017] The substrate processing apparatus 100, the database storage device 200, the information processing apparatus 300, and the learning apparatus 400 are connected to a network 501 and can transmit and receive data to and from each other. The network 501 may be, for example, a local area network (LAN) or a wide area network (WAN). Alternatively, the network 501 may be the Internet. The connection form of the network 501 may be a wired connection or a wireless connection. Further, the substrate processing apparatus 100 and the information processing apparatus 300 may be connected by a dedicated network instead of the network 501.

[0018] The substrate processing apparatus 100 executes a series of processes of forming a solid film or a liquid film on a substrate by supplying a processing liquid to the substrate under predetermined processing conditions, and then removing the solid film or the liquid film from the substrate to dry the substrate. Specifically, the substrate processing apparatus 100 applies a processing liquid, which is a mixed solution of a sublimating agent dissolved in a solvent at a predetermined concentration, to the surface of the substrate on which a pattern is formed, and then evaporates the solvent while rotating the substrate to deposit the sublimating agent on the substrate. Thereafter, the substrate processing apparatus 100 blows an inert gas onto the substrate for sublimation drying.

[0019] The database storage device 200 includes a large-capacity storage device such as a server. In the database storage device 200, a reference sublimating agent database indicating molecular descriptors (hereinafter simply referred to as descriptors) and physical property values of a reference sublimating agent is stored in advance. The reference sublimating agent is a general sublimating agent that is not necessarily used for sublimation drying of the substrate. Further, in the database storage device 200, a reference solvent database indicating descriptors and physical property values of a reference solvent is stored in advance. The reference solvent is a general solvent that is not necessarily used for sublimation drying of the substrate. Furthermore, in the database storage device 200, a result database indicating the relationship between the combination of the sublimating agent and the solvent used for the actual sublimation drying of the substrate, the combination of a plurality of processing conditions of the substrate, and the processing result is stored.

[0020] The information processing apparatus 300 is configured by, for example, a general-purpose computer. FIG. 2 is a diagram showing an example of the configuration of the information processing apparatus 300. As shown in FIG. 2, the information processing apparatus 300 includes a CPU (Central Processing Unit) 310, a RAM (Random Access Memory) 320, a ROM (Read Only Memory) 330, a storage device 340, an operation unit 350, a display device 360, an input / output I / F (Interface) 370, and a bus 380. The CPU 310, the RAM 320, the ROM 330, the storage device 340, the operation unit 350, the display device 360, and the input / output I / F 370 are connected to the bus 380.

[0021] The RAM 320 is used as a working area for the CPU 310. A system program is stored in the ROM 330. The storage device 340 includes a storage medium such as a hard disk or a semiconductor memory, and stores a recipe determination program for executing the recipe determination process described later. The recipe determination program may be stored in the ROM 330 or another external storage device.

[0022] The operation unit 350 is an input device such as a keyboard, a mouse, or a touch panel. The user can give a predetermined instruction to the information processing apparatus 300 by operating the operation unit 350. The display device 360 is a display device such as a liquid crystal display device, and displays a GUI (Graphical User Interface) or the like for receiving an instruction from the user. The input / output I / F 370 is connected to the network 501.

[0023] The learning device 400 is configured by, for example, a general-purpose computer. Therefore, since the learning device 400 has basically the same configuration as the information processing device 300 except for the following points, the detailed configuration of the learning device 400 will be omitted. The storage device of the learning device 400 stores a learning program for executing the learning process described later. The learning program may be stored in the ROM of the learning device 400 or other external storage devices. Note that the information processing device 300 and the learning device 400 may be configured by the same hardware. In this case, the programs installed in the information processing device 300 and the learning device 400 are different. Therefore, the processes executed by the information processing device 300 and the learning device 400 are different from each other.

[0024] 2. Database Storage Device FIG. 3 is a diagram showing an example of a reference sublimating agent database stored in advance in the database storage device 200. As shown in FIG. 3, in the reference sublimating agent database 201, a plurality of descriptors and a plurality of physical property values are shown for each of the plurality of reference sublimating agents. In the reference sublimating agent database 201, there are no missing descriptors and physical property values for each reference sublimating agent.

[0025] Specific examples of the descriptors of the reference sublimating agent include RDKit, Mordred, or fingerprints (such as Morgan or MACCS Keys). Specific examples of the physical property values of the sublimating agent include crystal structure, lattice constant, vapor pressure, molecular weight, density, molar volume, melting point, glass transition temperature, boiling point, heat of evaporation, heat of fusion, heat of sublimation, viscosity, HSP (Hansen solubility parameter), chemical potential, hydration energy, partition coefficient, interfacial free energy with the substrate, refractive index, dielectric constant, polarizability, dipole moment, HOMO / LUMO, or HOMO / LUMO gap.

[0026] FIG. 4 is a diagram showing an example of a reference solvent database stored in advance in the database storage device 200. As shown in FIG. 4, in the reference solvent database 202, a plurality of descriptors and a plurality of physical property values are shown for each of the plurality of reference solvents. In the reference solvent database 202, there are no missing descriptors and physical property values for each reference solvent.

[0027] Specific examples of the descriptors of the reference solvent include RDKit, Mordred, or fingerprints (such as Morgan or MACCS Keys). Specific examples of the physical property values of the reference solvent include vapor pressure, molecular weight, density, molar volume, melting point, glass transition temperature, boiling point, heat of vaporization, heat of fusion, heat of sublimation, viscosity, HSP, chemical potential, hydration energy, partition coefficient, surface tension, interfacial free energy with the substrate, refractive index, dielectric constant, polarizability, dipole moment, HOMO / LUMO, or HOMO / LUMO gap.

[0028] FIG. 5 is a diagram showing an example of a result database stored in the database storage device 200. As shown in FIG. 5, in the result database 203, the relationship between the combination of the sublimating agent and the solvent used for the sublimation drying of the actual substrate, the combination of a plurality of processing conditions of the substrate, and the processing result is shown. In this example, the processing result is the collapse rate of the pattern formed on the surface of the substrate, and is evaluated, for example, by image analysis of the substrate.

[0029] Specific examples of the processing conditions of the substrate include the rotation speed of the substrate, the temperature of the solution, the temperature of the substrate, the temperature inside the chamber described later, the humidity inside the chamber, the processing time, the flow rate of the inert gas, the flow rate of the processing liquid, the air flow inside the chamber, the discharge position of each nozzle described later, the residence time of each nozzle, the swinging behavior of each nozzle, or the pressure inside the chamber.

[0030] The result database 203 may be updated by adding new data each time the sublimation drying of the substrate is performed. In this case, the data used for sublimation drying and the processing results may be added to the result database 203 by the user. Alternatively, some data such as the sublimating agent, solvent, or processing conditions may be directly added from the substrate processing apparatus 100 of FIG. 1 to the result database 203.

[0031] 3. Learning device FIG. 6 is a block diagram showing the functional configuration of the learning device 400 of FIG. 1. As shown in FIG. 6, the learning device 400 includes, as functional units, a reference sublimating agent acquisition unit 401, a sublimating agent model generation unit 402, a reference solvent acquisition unit 403, a solvent model generation unit 404, a result acquisition unit 405, a sublimating agent identification unit 406, a solvent identification unit 407, a sublimating agent completion unit 408, a solvent completion unit 409, a database generation unit 410, a processing liquid calculation unit 411, a result model generation unit 412, and an optimization unit 413. When the CPU of the learning device 400 executes a learning program, the functional units of the learning device 400 are realized. Some or all of the functional units of the learning device 400 may be realized by hardware such as an electronic circuit.

[0032] The reference sublimating agent acquisition unit 401 acquires the reference sublimating agent database 201 of FIG. 3 from the database storage device 200. The reference sublimating agent acquisition unit 401 may acquire the reference sublimating agent database 201 from a server or the like provided outside the substrate processing system 500. The sublimating agent model generation unit 402 trains a predetermined machine learning model using the descriptors of the reference sublimating agent in the reference sublimating agent database 201 acquired by the reference sublimating agent acquisition unit 401 as explanatory variables and the physical property values of the reference sublimating agent as objective variables, thereby generating a sublimating agent model for predicting the physical property values of the sublimating agent from the descriptors of the sublimating agent.

[0033] The reference solvent acquisition unit 403 acquires the reference solvent database 202 in FIG. 4 from the database storage device 200. The reference solvent acquisition unit 403 may acquire the reference solvent database 202 from a server or the like provided outside the substrate processing system 500. The solvent model generation unit 404 trains a predetermined machine learning model using the descriptors of the reference solvents in the reference solvent database 202 acquired by the reference solvent acquisition unit 403 as explanatory variables and the physical property values of the reference solvents as objective variables, thereby generating a solvent model for predicting the physical property values of solvents from the descriptors of the solvents.

[0034] The result acquisition unit 405 acquires the result database 203 in FIG. 5 from the database storage device 200. The sublimating agent identification unit 406 identifies the sublimating agent used for the sublimation drying of the actual substrate based on the result database 203 acquired by the result acquisition unit 405, and also identifies the descriptors and physical property values of the sublimating agent. In this example, the sublimating agent identification unit 406 generates a processing-use sublimating agent database showing the descriptors and physical property values of the identified sublimating agent.

[0035] FIG. 7 is a diagram showing an example of the processing-use sublimating agent database generated by the sublimating agent identification unit 406. As shown in FIG. 7, the processing-use sublimating agent database 204 shows the descriptors and physical property values for each of a plurality of sublimating agents used for the sublimation drying of the actual substrate. The processing-use sublimating agent database 204 may be generated by the user operating the operation unit of the learning device 400 to input the descriptors and physical property values of the sublimating agent used for the sublimation drying of the actual substrate to the sublimating agent identification unit 406. Here, the physical property values of the sublimating agent used for the sublimation drying of the actual substrate are often not investigated. Therefore, in the processing-use sublimating agent database 204, there may be a lack of physical property values for each sublimating agent.

[0036] The solvent identification unit 407 identifies the solvent used for the sublimation drying of the actual substrate based on the result database 203 acquired by the result acquisition unit 405, and also identifies the descriptors and physical property values of the solvent. In this example, the solvent identification unit 407 generates a processing-use solvent database showing the descriptors and physical property values of the identified solvent.

[0037] FIG. 8 is a diagram showing an example of the processing solvent database generated by the solvent specifying unit 407. As shown in FIG. 8, the processing solvent database 205 shows descriptors and physical property values for each of a plurality of solvents used for the sublimation drying of an actual substrate. The processing solvent database 205 may be generated by a user operating the operation unit of the learning device 400 to input the descriptors and physical property values of the solvents used for the sublimation drying of the actual substrate into the solvent specifying unit 407. Here, the physical property values of the solvents used for the sublimation drying of the actual substrate are often not investigated. Therefore, in the processing solvent database 205, there may be a lack of physical property values for each solvent.

[0038] The sublimating agent complementing unit 408 uses the sublimating agent model generated by the sublimating agent model generating unit 402 to predict the physical property values of each sublimating agent from the descriptors of the processing sublimating agent database 204 generated by the sublimating agent specifying unit 406. Further, the sublimating agent complementing unit 408 uses the predicted physical property values of the sublimating agent to complement the missing physical property values of the processing sublimating agent database 204. In this example, the sublimating agent complementing unit 408 generates a complemented sublimating agent database in which the physical property values of the sublimating agent are complemented. FIG. 9 is a diagram showing an example of the complemented sublimating agent database generated by the sublimating agent complementing unit 408. As shown in FIG. 9, in the complemented sublimating agent database 206, there are no missing descriptors and physical property values for each sublimating agent.

[0039] The solvent complementing unit 409 uses the solvent model generated by the solvent model generating unit 404 to predict the physical property values of each solvent from the descriptors of the processing solvent database 205 generated by the solvent specifying unit 407. Further, the solvent complementing unit 409 uses the predicted physical property values of the solvent to complement the missing physical property values of the processing solvent database 205. In this example, the solvent complementing unit 409 generates a complemented solvent database in which the physical property values of the solvent are complemented. FIG. 10 is a diagram showing an example of the complemented solvent database generated by the solvent complementing unit 409. As shown in FIG. 10, in the complemented solvent database 207, there are no missing descriptors and physical property values for each solvent.

[0040] The database generation unit 410 generates an integrated database. FIG. 11 is a diagram showing an example of the integrated database generated by the database generation unit 410. As shown in FIG. 11, in the integrated database 208, the result database 203 of FIG. 5 acquired by the result acquisition unit 405, the complementary sublimating agent database 206 of FIG. 9 generated by the sublimating agent complementary unit 408, and the complementary solvent database 207 of FIG. 10 generated by the solvent complementary unit 409 are integrated.

[0041] The treatment liquid calculation unit 411 calculates the descriptors and physical property values of the treatment liquid in which the sublimating agent is dissolved in the solvent from the descriptors and physical property values of the sublimating agent and the solvent in the integrated database 208. In this example, the treatment liquid calculation unit 411 generates a learning database using the calculated descriptors and physical property values of the treatment liquid. FIG. 12 is a diagram showing an example of the learning database generated by the treatment liquid calculation unit 411. As shown in FIG. 12, in the learning database 209, the descriptors and physical property values of the treatment liquid calculated by the treatment liquid calculation unit 411 are added to the integrated database 208.

[0042] Specific examples of the descriptors of the treatment liquid include the results of the four arithmetic operations of the same type of descriptors of the sublimating agent and the solvent. Specific examples of the physical property values of the treatment liquid include the results of the four arithmetic operations of the same type of physical property values of the sublimating agent and the solvent. In these four arithmetic operations, weights may be assigned to the descriptors or physical property values of the sublimating agent or the solvent based on the concentration ratio of the sublimating agent and the solvent. Further, other specific examples of the physical property values of the treatment liquid include the solubility of the sublimating agent in the solvent, the χ parameter or HSP of the sublimating agent in the solvent, the interfacial free energy between the solvent and the sublimating agent, or the heat of dissolution when the sublimating agent is dissolved in the solvent.

[0043] Based on the learning database 209 generated by the database generation unit 410, the result model generation unit 412 generates a result model for predicting the processing result of the substrate from the combination of the sublimating agent and the solvent and the processing conditions of the substrate. Specifically, the result model is generated by training a predetermined machine learning model with the descriptors and physical property values of the sublimating agent, solvent, and processing solution in the learning database 209 and the processing conditions of the substrate as explanatory variables and the processing conditions of the substrate as the objective variable.

[0044] Before the result model is generated, the optimization unit 413 optimizes the explanatory variables used for generating the result model. Specifically, the optimization unit 413 calculates the importance of the explanatory variables. As a method for calculating the importance, the SHAP (SHapley Additive exPlanations) value or the importance of the features of the extra tree may be calculated. In this case, the optimized result model is generated by the result model generation unit 412. Alternatively, after the result model is generated, the optimization unit 413 may optimize the explanatory variables and cause the result model generation unit 412 to relearn using the optimized explanatory variables. In this case, the result model generated by the result model generation unit 412 is updated to a result model that can predict the processing result with higher accuracy.

[0045] 4. Information Processing Apparatus FIG. 13 is a block diagram showing the functional configuration of the information processing apparatus 300 in FIG. 1. As shown in FIG. 13, the information processing apparatus 300 includes, as functional units, a model acquisition unit 301, a sublimating agent determination unit 302, a solvent determination unit 303, a condition selection unit 304, a result prediction unit 305, an evaluation unit 306, and a presentation unit 307. When the CPU 310 in FIG. 2 of the information processing apparatus 300 executes the recipe determination program, the functional units of the information processing apparatus 300 are realized. A part or all of the functional units of the information processing apparatus 300 may be realized by hardware such as an electronic circuit.

[0046] The model acquisition unit 301 acquires the result model generated by the result model generation unit 412 in FIG. 6 of the learning device 400. The sublimating agent determination unit 302 determines a sublimating agent to be used for the sublimation drying of the substrate based on the user's designation. The solvent determination unit 303 determines a solvent to be used for the sublimation drying of the substrate based on the user's designation. The user can respectively specify an appropriate sublimating agent and solvent for the substrate to be processed to the sublimating agent determination unit 302 and the solvent determination unit 303 by operating the operation unit 350 in FIG. 2.

[0047] The condition selection unit 304 selects processing conditions for performing a series of processes on the substrate from among a plurality of processing conditions. The processing conditions include combinations of a plurality of conditions such as the rotation speed of the substrate, the temperature of the solution, and the flow rate of the inert gas in a series of processes. The selection of the processing conditions is performed using a known method such as Bayesian optimization, the steepest descent method, or a genetic algorithm. The result prediction unit 305 uses the result model acquired by the model acquisition unit 301 to predict the processing result of the substrate from the sublimating agent determined by the sublimating agent determination unit 302, the solvent determined by the solvent determination unit 303, and the processing conditions selected by the condition selection unit 304.

[0048] The evaluation unit 306 evaluates a score indicating the goodness of the processing result of the substrate predicted by the result prediction unit 305. In this example, the lower the collapse rate of the pattern, the higher the score is given. Further, the evaluation unit 306 specifies the processing conditions selected by the condition selection unit 304 when the processing result with the highest score is obtained. When one or more processing results with a score equal to or higher than a predetermined threshold are obtained, the evaluation unit 306 may specify one or more processing conditions respectively selected by the condition selection unit 304 when the one or more processing results are obtained.

[0049] The presentation unit 307 presents the processing conditions specified by the evaluation unit 306 to the user. The presentation of the processing conditions may be performed, for example, by displaying the processing conditions on the display device 360 in FIG. 2. Thereby, the user can set the optimal processing conditions for performing sublimation drying of the substrate in the substrate processing apparatus 100 in FIG. 1 by visually recognizing the processing conditions displayed on the display device 360.

[0050] 5. Substrate Processing Apparatus FIG. 14 is a schematic view of the inside of the substrate processing apparatus 100 seen horizontally. As shown in FIG. 14, the substrate processing apparatus 100 includes a processing unit 2 and a control device 3. The control device 3 can communicate with the information processing apparatus 300 of FIG. 1 and controls the processing unit 2 to sublimation-dry the substrate W under the processing conditions set by the user. The processing unit 2 includes a box-shaped chamber 4, a spin chuck 10 that rotates around a vertical rotation axis A1 passing through the central portion of the substrate W while holding one substrate W horizontally in the chamber 4, and a cylindrical processing cup 21 that surrounds the spin chuck 10 around the rotation axis A1.

[0051] The spin chuck 10 includes a disc-shaped spin base 12 held in a horizontal posture, a plurality of chuck pins 11 that hold the substrate W in a horizontal posture above the spin base 12, a spin shaft 13 that extends downward from the central portion of the spin base 12, and a spin motor 14 that rotates the spin base 12 and the plurality of chuck pins 11 by rotating the spin shaft 13.

[0052] The processing cup 21 includes a plurality of guards 24 that receive the processing liquid discharged outward from the substrate W, a plurality of cups 23 that receive the processing liquid guided downward by the plurality of guards 24, and a cylindrical outer wall member 22 that surrounds the plurality of guards 24 and the plurality of cups 23. The plurality of guards 24 can be individually raised and lowered by a guard lifting unit 27.

[0053] The processing unit 2 includes a chemical liquid nozzle 31 that discharges a chemical liquid, a rinse liquid nozzle 35 that discharges a rinse liquid, a processing liquid nozzle 39 that discharges a processing liquid, and a replacement liquid nozzle 43 that discharges a replacement liquid. The chemical liquid nozzle 31, the rinse liquid nozzle 35, the processing liquid nozzle 39, and the replacement liquid nozzle 43 can be independently moved horizontally in the chamber 4 by nozzle moving units 34, 38, 42, 46 provided corresponding to each of them.

[0054] The processing liquid nozzle 39 is connected to a processing liquid pipe 40 that guides the processing liquid to the processing liquid nozzle 39. When a processing liquid valve 41 interposed in the processing liquid pipe 40 is opened, the processing liquid is continuously discharged downward from the discharge port of the processing liquid nozzle 39. The processing liquid is a mixed liquid containing a sublimating agent and a solvent that melts with the sublimating agent.

[0055] The processing liquid nozzle 39 is connected to a nozzle moving unit 42. The nozzle moving unit 42 moves the processing liquid nozzle 39 in at least one of the vertical direction and the horizontal direction. The nozzle moving unit 42 horizontally moves the processing liquid nozzle 39 between a processing position where the processing liquid discharged from the processing liquid nozzle 39 is supplied to the upper surface of the substrate W and a standby position where the processing liquid nozzle 39 is positioned around the processing cup 21 in plan view.

[0056] The processing unit 2 includes a disk-shaped blocking member 51 disposed above the spin chuck 10. The blocking member 51 includes a disk portion 52 horizontally disposed above the spin chuck 10. The blocking member 51 is horizontally supported by a cylindrical support shaft 53 extending upward from the central portion of the disk portion 52. The center line of the disk portion 52 is disposed on the rotation axis A1 of the substrate W. The lower surface of the disk portion 52 corresponds to the lower surface 51L of the blocking member 51. The lower surface 51L of the blocking member 51 is parallel to the upper surface of the substrate W and has an outer diameter equal to or larger than the diameter of the substrate W.

[0057] The blocking member 51 is connected to a blocking member lifting unit 54 that vertically moves the blocking member 51. The blocking member lifting unit 54 moves the blocking member 51 to an arbitrary position from the upper position (the position shown in FIG. 14) to the lower position.

[0058] A central nozzle 55 is disposed in a through hole that vertically penetrates the central portion of the blocking member 51. The central nozzle 55 moves up and down together with the blocking member 51. The central nozzle 55 is connected to an upper gas pipe 56 that guides an inert gas to the central nozzle 55. The substrate processing apparatus 100 includes an upper temperature regulator 59 that heats or cools the inert gas discharged from the central nozzle 55. When an upper gas valve 57 interposed in the upper gas pipe 56 is opened, the inert gas is continuously discharged downward from the discharge port of the central nozzle 55 at a flow rate corresponding to the opening degree of a flow rate adjustment valve 58 that changes the flow rate of the inert gas. The inert gas discharged from the central nozzle 55 is nitrogen gas.

[0059] 6. Learning Process FIGS. 15 and 16 are flowcharts showing an example of the flow of the learning process. The learning process is a process executed by a CPU included in the learning apparatus 400 by executing a learning program. Hereinafter, the learning process of FIGS. 15 and 16 will be described with reference to the learning apparatus 400 of FIG. 6 and various databases of FIGS. 3 to 5 and FIGS. 7 to 12.

[0060] First, a reference sublimating agent acquisition unit 401 acquires a reference sublimating agent database 201 of FIG. 3 from the database storage device 200 (step S1). Next, a sublimating agent model generation unit 402 generates a sublimating agent model that predicts physical property values of the sublimating agent from descriptors of the sublimating agent based on the reference sublimating agent database 201 acquired in step S1 (step S2).

[0061] Also, a reference solvent acquisition unit 403 acquires a reference solvent database 202 of FIG. 4 from the database storage device 200 (step S3). Next, a solvent model generation unit 404 generates a solvent model that predicts physical property values of the solvent from descriptors of the solvent based on the reference solvent database 202 acquired in step S3 (step S4).

[0062] Further, the result acquisition unit 405 acquires the result database 203 in FIG. 5 from the database storage device 200 (step S5). Next, the sublimating agent identification unit 406 identifies the sublimating agent used for sublimation drying, the descriptor of the sublimating agent, and the physical property values based on the result database 203 acquired in step S5 (step S6). Subsequently, the sublimating agent identification unit 406 generates the sublimating agent database 204 for processing in FIG. 7 based on the sublimating agent, the descriptor, and the physical property values identified in step S6 (step S7).

[0063] Also, the solvent identification unit 407 identifies the solvent used for sublimation drying, the descriptor of the solvent, and the physical property values based on the result database 203 acquired in step S5 (step S8). Subsequently, the solvent identification unit 407 generates the solvent database 205 for processing in FIG. 8 based on the solvent, the descriptor, and the physical property values identified in step S8 (step S9). Either of step S6, S7 and step S8, S9 may be executed first, or they may be executed simultaneously. Also, either of step S1, S2, step S3, S4, and step S5 to S9 may be executed first, or they may be executed simultaneously.

[0064] Next, the sublimating agent complementing unit 408 predicts the physical property values of each sublimating agent in the sublimating agent database 204 for processing generated in step S7 from the descriptors of the sublimating agents using the sublimating agent model generated in step S2 (step S10). Subsequently, the sublimating agent complementing unit 408 complements the missing physical property values in the sublimating agent database 204 for processing using the physical property values of the sublimating agents predicted in step S10 (step S11). Thereafter, the sublimating agent complementing unit 408 generates the complemented sublimating agent database 206 in FIG. 9 in which the physical property values of the sublimating agents are complemented (step S12).

[0065] Further, the solvent completion unit 409 uses the solvent model generated in step S4 to predict the physical property values of each solvent from the descriptors of the solvent in the processing solvent database 205 generated in step S9 (step S13). Subsequently, the solvent completion unit 409 uses the physical property values of the solvent predicted in step S13 to complete the missing physical property values in the processing solvent database 205 (step S14). Thereafter, the solvent completion unit 409 generates the completed solvent database 207 shown in FIG. 10 in which the physical property values of the solvent are completed (step S15). Either step S10 to S12 or step S13 to S15 may be executed first, or they may be executed simultaneously.

[0066] Next, the database generation unit 410 generates the integrated database 208 shown in FIG. 11 by integrating the completed sublimating agent database 206 generated in step S12 and the completed solvent database 207 generated in step S13 (step S16). Subsequently, the processing liquid calculation unit 411 calculates the descriptors and physical property values of the processing liquid in which the sublimating agent is dissolved in the solvent based on the integrated database 208 generated in step S16 (step S17). Thereafter, the processing liquid calculation unit 411 generates the learning database 209 shown in FIG. 12 using the descriptors and physical property values of the processing liquid calculated in step S17 (step S18).

[0067] Next, the optimization unit 413 optimizes the explanatory variables in the learning database 209 generated in step S18 (step S19). The explanatory variables include the descriptors and physical property values of each of the sublimating agent, the solvent, and the processing liquid, and the processing conditions of the substrate. Next, the result model generation unit 412 generates a result model based on the learning database 209 generated in step S18 and in which the explanatory variables are optimized in step S19 (step S20). Thus, the learning process ends. Instead of or in addition to optimizing the explanatory variables in step S19, the optimization unit 413 may cause the result model generation unit 412 to perform re-learning using the optimized explanatory variables after step S20.

[0068] 7. Recipe determination process FIG. 17 is a flowchart showing an example of the flow of recipe determination processing. The recipe determination processing is a process executed by the CPU 310 shown in FIG. 2 included in the information processing apparatus 300 when the CPU 310 executes a recipe determination program. Hereinafter, the recipe determination processing in FIG. 17 will be described with reference to the information processing apparatus 300 in FIG. 13.

[0069] First, the model acquisition unit 301 acquires the result model generated in step S20 of the learning process (step S31). Further, the sublimating agent determination unit 302 determines a sublimating agent to be used for sublimation drying of the substrate W based on the designation of the user (step S32). The solvent determination unit 303 determines a solvent to be used for sublimation drying of the substrate based on the designation of the user (step S33). Any of steps S31 to S34 may be executed first or may be executed simultaneously.

[0070] Next, the condition selection unit 304 selects processing conditions for performing a series of processes on the substrate W from among a plurality of processing conditions (step S34). Subsequently, the result prediction unit 305 predicts the processing result of the substrate W from the sublimating agent determined in step S32, the solvent determined in step S33, and the processing conditions selected in step S34 using the result model acquired in step S31 (step S35). Thereafter, the evaluation unit 306 evaluates a score indicating the goodness of the processing result of the substrate W predicted in step S35 (step S36).

[0071] Next, the evaluation unit 306 determines whether the processing conditions have been searched a predetermined number of times (step S37). If the processing conditions have not been searched the predetermined number of times, the process returns to step S34. In this case, in step S34, the condition selection unit 304 selects the processing conditions again. Here, as described above, the selection of the processing conditions is performed using Bayesian optimization, the steepest descent method, a genetic algorithm, or the like, so that the selection of the optimal processing conditions can be performed in a relatively short time. Thereafter, steps S35 to S37 are sequentially executed.

[0072] In step S37, when the processing conditions are searched for a determined number of times, appropriate processing conditions are specified based on the score evaluated in step S36 (step S38). In step S38, the processing conditions when the processing result having the highest score is obtained may be specified as the appropriate processing conditions. Alternatively, the processing conditions when a processing result having a score equal to or higher than a threshold value is obtained may be specified as the appropriate processing conditions.

[0073] Next, the presentation unit 307 presents the processing conditions specified in step S38 (step S39), and ends the recipe determination process. In step S39, the specified processing conditions may be displayed on the display device 360 in FIG. 2. Further, when a plurality of processing conditions are specified in step S38, the processing conditions may be displayed on the display device 360 in descending order of the scores of the processing results. In this case, the user can easily recognize the processing conditions that are more likely to be appropriate.

[0074] 8. Substrate Processing FIG. 18 is a process diagram for explaining an example of substrate processing by the substrate processing apparatus 100. Hereinafter, the substrate processing in FIG. 18 will be described with reference to the substrate processing apparatus 100 in FIG. 14. When the substrate W is processed by the substrate processing apparatus 100, the substrate W is carried into the chamber 4 and held by the spin chuck 10. Thereafter, the guard lifting unit 27 is in a state where at least one guard 24 is lifted from the lower position to the upper position. The surface of the substrate W is a surface on which devices such as transistors and capacitors are formed, and a pattern is formed.

[0075] First, the spin motor 14 is driven to start the rotation of the substrate W (step S41). Next, a chemical solution supply step is performed in which a chemical solution is supplied onto the upper surface of the substrate W to form a liquid film of the chemical solution that covers the entire upper surface of the substrate W (step S42). Specifically, the chemical solution nozzle 31 is moved from the standby position to the processing position, the chemical solution nozzle 31 discharges the chemical solution for a predetermined time, and then moves to the standby position. The chemical solution discharged from the chemical solution nozzle 31 collides with the upper surface of the rotating substrate W and then flows outward along the upper surface of the substrate W by centrifugal force.

[0076] Next, a rinse liquid supply step of flushing the chemical solution on the substrate W is performed by supplying pure water, which is an example of the rinse liquid, to the upper surface of the substrate W (step S43). Specifically, the rinse liquid nozzle 35 moves from the standby position to the processing position, the rinse liquid nozzle 35 discharges the rinse liquid for a predetermined time, and then moves to the standby position. The pure water discharged from the rinse liquid nozzle 35 collides with the upper surface of the rotating substrate W and then flows outward along the upper surface of the substrate W by centrifugal force.

[0077] Next, a replacement liquid supply step of replacing the pure water on the substrate W with a replacement liquid is performed by supplying a replacement liquid that dissolves in both the rinse liquid and the processing liquid to the upper surface of the substrate W (step S44). Specifically, the replacement liquid nozzle 43 moves from the standby position to the processing position, the replacement liquid nozzle 43 discharges the replacement liquid for a predetermined time, and then moves to the standby position. The replacement liquid discharged from the replacement liquid nozzle 43 collides with the upper surface of the substrate W and then flows outward along the upper surface of the substrate W by centrifugal force. The pure water on the substrate W is replaced with the replacement liquid discharged from the replacement liquid nozzle 43. Thereby, a liquid film of the replacement liquid covering the entire upper surface of the substrate W is formed.

[0078] Next, a processing liquid supply step of forming a liquid film of the processing liquid on the substrate W is performed by supplying the processing liquid to the upper surface of the substrate W (step S45). Specifically, the nozzle moving unit 42 moves the processing liquid nozzle 39 from the standby position to the processing position. Thereafter, the processing liquid valve 41 is opened, and the processing liquid nozzle 39 starts discharging the processing liquid. Before the discharge of the processing liquid is started, the guard lifting unit 27 may vertically move at least one guard 24 to switch the guard 24 for receiving the liquid discharged from the substrate W. When a predetermined time has elapsed since the processing liquid valve 41 was opened, the processing liquid valve 41 is closed, and the discharge of the processing liquid is stopped. Thereafter, the nozzle moving unit 42 moves the processing liquid nozzle 39 to the standby position.

[0079] The processing liquid discharged from the processing liquid nozzle 39 collides with the upper surface of the substrate W rotating at the liquid supply rate, and then flows outward along the upper surface of the substrate W by centrifugal force. The replacement liquid on the substrate W is replaced by the processing liquid discharged from the processing liquid nozzle 39. Thereby, a liquid film of the processing liquid covering the entire upper surface of the substrate W is formed. When the processing liquid nozzle 39 discharges the processing liquid, the nozzle moving unit 42 moves the liquid landing position so that the liquid landing position of the processing liquid on the upper surface of the substrate W passes through the central portion and the outer peripheral portion.

[0080] Next, a film thickness reduction step is performed (step S46) to reduce the film thickness (thickness of the liquid film) of the processing liquid on the substrate W while maintaining the state where the entire upper surface of the substrate W is covered with the liquid film of the processing liquid by removing a part of the processing liquid on the substrate W. Specifically, with the blocking member 51 positioned at the lower position, the spin motor 14 rotates the substrate W. The processing liquid on the substrate W is discharged outward from the substrate W by centrifugal force even after the discharge of the processing liquid is stopped. Therefore, the thickness of the liquid film of the processing liquid on the substrate W decreases. When a certain amount of the processing liquid on the substrate W is discharged, the discharge amount of the processing liquid from the substrate W per unit time decreases to zero or substantially zero.

[0081] Next, a solid body forming step is performed (step S47) to form a solid body containing a sublimating agent on the substrate W by solidifying the processing liquid on the substrate W. Specifically, with the blocking member 51 positioned at the lower position, the spin motor 14 rotates the substrate W. Further, the upper gas valve 57 is opened to start discharging nitrogen gas from the central nozzle 55. In the solid body forming step, the evaporation of the processing liquid is promoted, and a part of the processing liquid on the substrate W evaporates. Therefore, while the concentration of the sublimating agent gradually increases, the film thickness of the processing liquid gradually decreases.

[0082] Next, a sublimation process is performed to remove the solidified material on the substrate W from the upper surface of the substrate W (step S48). Specifically, with the blocking member 51 in the lower position, the spin motor 14 rotates the substrate W. Further, the upper gas valve 57 is opened to start discharging nitrogen gas from the central nozzle 55. When the sublimation process is completed, the spin motor 14 stops and the rotation of the substrate W is stopped (step S49). Thereby, the substrate processing is completed.

[0083] 9. Effects In the learning device 400 according to the present embodiment, the result acquisition unit 405, the sublimating agent specifying unit 406, and the solvent specifying unit 407 acquire the result database 203, the sublimating agent database for processing 204, and the solvent database for processing 205, respectively. That is, data indicating the relationship between the processing conditions and at least one descriptor of the sublimating agent and the solvent and the processing result of the substrate W is acquired.

[0084] Here, in the sublimating agent or solvent used for the sublimation drying of the substrate W, physical property values are often not investigated. Therefore, the physical property values of the sublimating agent or solvent in the data may be missing. Even in such a case, the physical property values are predicted by the sublimating agent complementing unit 408 or the solvent complementing unit 409 from at least one descriptor of the sublimating agent and the solvent in the data, and the missing physical property values are complemented.

[0085] According to this configuration, the result model generation unit 412 can generate a result model for predicting the processing result of the substrate W by performing machine learning on the data with the physical property values complemented. Thereby, it is possible to search for optimal processing conditions based on the processing result of the substrate W. Therefore, even when there are missing physical property values of the sublimating agent or solvent, it is possible to generate a result model for searching for the optimal processing conditions of the substrate W.

[0086] As physical property values of the sublimating agent used for the sublimation drying of the substrate W, the vapor pressure, melting point, heat of dissolution, or Hansen solubility parameter are often lacking. Therefore, the physical property values of the sublimating agent supplemented by the sublimating agent supplementing unit 408 include at least one of the vapor pressure, melting point, heat of dissolution, and Hansen solubility parameter. Thereby, a result model for searching for the optimum processing conditions of the substrate W can be easily generated.

[0087] As physical property values of the solvent used for the sublimation drying of the substrate W, the vapor pressure, Hansen solubility parameter, viscosity, or surface tension are often lacking. Therefore, the physical property values of the solvent supplemented by the solvent supplementing unit 409 include at least one of the vapor pressure, Hansen solubility parameter, viscosity, and surface tension. Thereby, a result model for searching for the optimum processing conditions of the substrate W can be easily generated.

[0088] The processing liquid calculation unit 411 calculates the descriptors and physical property values of the processing liquid based on the descriptors and physical property values of the sublimating agent and the descriptors and physical property values of the solvent. The result model predicts the processing result of the substrate W based on the descriptors and physical property values of the processing liquid calculated by the processing liquid calculation unit 411. In this case, based on the descriptors and physical property values of the processing liquid, a result model for searching for the optimum processing conditions of the substrate W can be generated more appropriately.

[0089] As physical property values of the processing liquid used for the sublimation drying of the substrate W, the solubility of the sublimating agent in the solvent is often lacking. Therefore, the physical property values of the processing liquid include the solubility of the sublimating agent in the solvent. Thereby, a result model for searching for the optimum processing conditions of the substrate W can be easily generated.

[0090] The result model generation unit 412 may perform machine learning on the processing result of the substrate W, using at least a part of the processing conditions, the descriptor and physical property values of the sublimating agent, and the descriptor and physical property values of the solvent as explanatory variables. In this case, a result model for predicting the processing result of the substrate W can be easily generated by machine learning. In this example, the explanatory variables are optimized by the optimization unit 413 based on the importance of the explanatory variables. The result model generation unit 412 performs machine learning on the explanatory variables optimized by the optimization unit 413. Therefore, the result model can be generated efficiently.

[0091] The sublimating agent model generation unit 402 generates a sublimating agent model for predicting physical property values by performing machine learning on the relationship between the descriptor and physical property values of the reference sublimating agent in the reference sublimating agent database 201. The sublimating agent complementing unit 408 predicts physical property values from the descriptor of the sublimating agent using the sublimating agent model generated by the sublimating agent model generation unit 402, and complements the missing physical property values of the sublimating agent in the processing-use sublimating agent database 204 obtained by the sublimating agent specifying unit 406 using the predicted physical property values. In this case, based on the relationship between the descriptor and physical property values of the reference sublimating agent, the physical property values of the sublimating agent used for the sublimation drying of the substrate W can be easily predicted, and the missing physical property values can be complemented.

[0092] The solvent model generation unit 404 generates a solvent model for predicting physical property values by performing machine learning on the relationship between the descriptor and physical property values of the reference solvent in the reference solvent database 202. The solvent complementing unit 409 predicts physical property values from the descriptor of the solvent using the solvent model generated by the solvent model generation unit 404, and complements the missing physical property values of the solvent in the processing-use solvent database 205 obtained by the solvent specifying unit 407 using the predicted physical property values. In this case, based on the relationship between the descriptor and physical property values of the reference solvent, the physical property values of the solvent used for the sublimation drying of the substrate W can be easily predicted, and the missing physical property values can be complemented.

[0093] In the information processing apparatus 300 according to the present embodiment, based on the result model generated by the learning apparatus 400, the processing result of the substrate W is predicted from the sublimating agent, the solvent, and the processing conditions used for the sublimation drying of the substrate W. Therefore, based on the predicted processing result of the substrate W, the optimum processing conditions for the substrate W can be searched. Here, the condition selection unit 304 selects the processing conditions using Bayesian optimization, the steepest descent method, or a genetic algorithm. In this case, the optimum processing conditions for the substrate W can be efficiently searched.

[0094] The evaluation unit 306 evaluates the quality of the processing result of the substrate W predicted by the result prediction unit 305. The presentation unit 307 presents the processing conditions corresponding to the processing result evaluated as good by the evaluation unit 306. In this case, the user can easily recognize the processing conditions when a good processing result is obtained. Thereby, the optimum processing conditions for the substrate W can be easily searched.

[0095] According to the substrate processing system 500 according to the present embodiment, based on the processing result of the substrate W predicted by the information processing apparatus 300, the substrate W can be sublimation-dried under the optimum processing conditions in the substrate processing apparatus 100.

[0096] 10. Other Embodiments (1) In the above embodiment, the learning apparatus 400 includes the processing liquid calculation unit 411 and the optimization unit 413, but the embodiment is not limited thereto. When the result model is generated without using the molecular descriptor and the physical property value of the processing liquid, the learning apparatus 400 may not include the processing liquid calculation unit 411. Further, when the optimization of the explanatory variables in machine learning is not performed, the learning apparatus 400 may not include the optimization unit 413.

[0097] (2) In the above-described embodiment, the sublimating agent specifying unit 406, the solvent specifying unit 407, the sublimating agent supplementing unit 408, the solvent supplementing unit 409, the database generating unit 410, and the processing liquid calculating unit 411 generate a database as appropriate, but the embodiment is not limited thereto. The sublimating agent specifying unit 406, the solvent specifying unit 407, the sublimating agent supplementing unit 408, the solvent supplementing unit 409, the database generating unit 410, or the processing liquid calculating unit 411 may not generate a database. In this case, since the integrated database 208 is not generated, the learning device 400 does not include the database generating unit 410.

[0098] (3) In the above-described embodiment, the information processing apparatus 300 includes the evaluation unit 306 and the presentation unit 307, but the embodiment is not limited thereto. When the user determines the quality of the processing result of the substrate W predicted by the result prediction unit 305, the information processing apparatus 300 may not include the evaluation unit 306 and the presentation unit 307.

[0099] (4) In the above-described embodiment, the condition selection unit 304 selects the processing conditions using Bayesian optimization, the steepest descent method, or a genetic algorithm, but the embodiment is not limited thereto. When the information processing apparatus 300 has a sufficient processing speed, the condition selection unit 304 may randomly select the processing conditions.

[0100] 11. Correspondence between each component of the claims and each part of the embodiment Hereinafter, examples of the correspondence between each component of the claims and each element of the embodiment will be described, but the present invention is not limited to the following examples. As each component of the claims, various other elements having the configurations or functions described in the claims can also be used.

[0101] In the above embodiment, the result acquisition unit 405, the sublimating agent identification unit 406, and the solvent identification unit 407 are examples of the data acquisition unit, the sublimating agent completion unit 408 and the solvent completion unit 409 are examples of the data completion unit, and the result model generation unit 412 is an example of the result model generation unit. The learning device 400 is an example of the learning device, the processing liquid calculation unit 411 is an example of the processing liquid calculation unit, the optimization unit 413 is an example of the optimization unit, the sublimating agent model generation unit 402 is an example of the sublimating agent model generation unit, and the sublimating agent completion unit 408 is an example of the sublimating agent completion unit.

[0102] The solvent model generation unit 404 is an example of the solvent model generation unit, the solvent completion unit 409 is an example of the solvent completion unit, the information processing device 300 is an example of the information processing device, the model acquisition unit 301 is an example of the model acquisition unit, and the sublimating agent determination unit 302 is an example of the sublimating agent determination unit. The solvent determination unit 303 is an example of the solvent determination unit, the condition selection unit 304 is an example of the condition selection unit, the result prediction unit 305 is an example of the result prediction unit, the evaluation unit 306 is an example of the evaluation unit, the presentation unit 307 is an example of the presentation unit, and the substrate processing system 500 is an example of the substrate processing system.

[0103] 12. Summary of the Embodiment (Item 1) The learning device according to Item 1 A data acquisition unit that acquires data indicating the relationship between the processing conditions in the process of sublimating and drying a substrate using a processing liquid containing a sublimating agent and a solvent and at least one molecular descriptor of the sublimating agent and the solvent, and the processing result of the substrate; A data completion unit that complements the lack of physical property values of at least one of the sublimating agent and the solvent in the data acquired by the data acquisition unit using a prediction algorithm that predicts physical property values from molecular descriptors; And a result model generation unit that generates a result model for predicting the processing result of the substrate by machine learning the data whose physical property values have been complemented by the data completion unit.

[0104] In this learning device, the data acquisition unit acquires data indicating the relationship between the processing conditions, at least one molecular descriptor of the sublimating agent and the solvent, and the processing result of the substrate. Here, in the sublimating agent or solvent used for the sublimation drying of the substrate, physical property values are often not examined. Therefore, the physical property values of the sublimating agent or solvent in the data may be missing. Even in such a case, the physical property values are predicted by the data complementation unit from at least one molecular descriptor of the sublimating agent and the solvent in the data, and the missing physical property values are complemented.

[0105] According to this configuration, the result model generation unit can generate a result model for predicting the processing result of the substrate by performing machine learning on the data with the physical property values complemented. Thereby, the optimal processing conditions can be searched based on the processing result of the substrate. Therefore, even when there are missing physical property values of the sublimating agent or solvent, it is possible to generate a result model for searching the optimal processing conditions of the substrate.

[0106] (Item 2) In the learning device according to Item 1, The physical property values of the sublimating agent complemented by the data complementation unit may include at least one of vapor pressure, melting point, heat of solution, and Hansen solubility parameter.

[0107] As the physical property values of the sublimating agent used for the sublimation drying of the substrate, vapor pressure, melting point, heat of solution, or Hansen solubility parameter is often missing. Even in this case, according to the above configuration, a result model for searching the optimal processing conditions of the substrate can be easily generated.

[0108] (Item 3) In the learning device according to Item 1 or Item 2, The physical property values of the solvent complemented by the data complementation unit may include at least one of vapor pressure, Hansen solubility parameter, viscosity, and surface tension.

[0109] As physical property values of the solvent used for sublimation drying of the substrate, vapor pressure, Hansen solubility parameter, viscosity, or surface tension are often lacking. Even in this case, according to the above configuration, a result model for searching for optimal processing conditions of the substrate can be easily generated.

[0110] (Item 4) The learning device according to any one of Items 1 to 3 further includes a treatment liquid calculation unit that calculates molecular descriptors and physical property values of the treatment liquid based on the molecular descriptors and physical property values of the sublimating agent and the molecular descriptors and physical property values of the solvent, The result model may predict the processing result of the substrate further based on the molecular descriptors and physical property values of the treatment liquid calculated by the treatment liquid calculation unit.

[0111] In this case, based on the molecular descriptors and physical property values of the treatment liquid, a result model for searching for optimal processing conditions of the substrate can be generated more appropriately.

[0112] (Item 5) In the learning device according to Item 4, The physical property value of the treatment liquid may include the solubility of the sublimating agent in the solvent.

[0113] As the physical property value of the treatment liquid used for sublimation drying of the substrate, the solubility of the sublimating agent in the solvent is often lacking. Even in this case, according to the above configuration, a result model for searching for optimal processing conditions of the substrate can be easily generated.

[0114] (Item 6) In the learning device according to any one of Items 1 to 5, The result model generation unit may perform machine learning using at least a part of processing conditions, molecular descriptors and physical property values of the sublimating agent, and molecular descriptors and physical property values of the solvent as explanatory variables and the processing result of the substrate as an objective variable.

[0115] In this case, a result model for predicting the processing result of the substrate by machine learning can be easily generated.

[0116] (Item 7) The learning device according to Item 6 further includes an optimization unit that optimizes the explanatory variables based on the importance of the explanatory variables, and the result model generation unit may perform machine learning on the explanatory variables optimized by the optimization unit.

[0117] In this case, a result model can be efficiently generated.

[0118] (Item 8) The learning device according to any one of Items 1 to 7 further includes a sublimating agent model generation unit that generates a sublimating agent model for predicting physical property values by performing machine learning on the relationship between the molecular descriptors and physical property values of a reference sublimating agent, and the data complementing unit may include a sublimating agent complementing unit that predicts physical property values from the molecular descriptors of the sublimating agent using the sublimating agent model generated by the sublimating agent model generation unit, and complements the missing physical property values of the sublimating agent in the data acquired by the data acquisition unit using the predicted physical property values.

[0119] In this case, based on the relationship between the molecular descriptors and physical property values of the reference sublimating agent, the physical property values of the sublimating agent used for sublimation drying of the substrate can be easily predicted, and the missing physical property values can be complemented.

[0120] (Item 9) The learning device according to any one of Items 1 to 8 further includes a solvent model generation unit that generates a solvent model for predicting physical property values by performing machine learning on the relationship between the molecular descriptors and physical property values of a reference solvent, and the data complementing unit may include a solvent complementing unit that predicts physical property values from the molecular descriptors of the solvent using the solvent model generated by the solvent model generation unit, and complements the missing physical property values of the solvent in the data acquired by the data acquisition unit using the predicted physical property values.

[0121] In this case, based on the relationship between the molecular descriptors and physical property values of the reference solvent, the physical property values of the solvent used for sublimation drying of the substrate can be easily predicted, and the missing physical property values can be complemented.

[0122] (Item 10) The information processing apparatus according to Item 10 is an information processing apparatus that uses a result model generated by the learning apparatus according to any one of Items 1 to 9, a model acquisition unit that acquires the result model; a sublimating agent determination unit that determines a sublimating agent used for sublimation drying of a substrate; a solvent determination unit that determines a solvent used for sublimation drying of a substrate; a condition selection unit that selects processing conditions; and a result prediction unit that predicts a processing result of a substrate from the sublimating agent determined by the sublimating agent determination unit, the solvent determined by the solvent determination unit, and the processing conditions selected by the condition selection unit, using the result model acquired by the model acquisition unit.

[0123] In this information processing apparatus, based on the result model generated by the learning apparatus, the processing result of the substrate is predicted from the sublimating agent, solvent, and processing conditions used for sublimation drying of the substrate. Therefore, based on the predicted processing result of the substrate, the optimal processing conditions of the substrate can be searched for.

[0124] (Item 11) In the information processing apparatus according to Item 10, the condition selection unit may select processing conditions using Bayesian optimization, the steepest descent method, or a genetic algorithm.

[0125] In this case, the optimal processing conditions of the substrate can be efficiently searched for.

[0126] (Item 12) The information processing apparatus according to Item 10 or 11 is further provided with an evaluation unit that evaluates the quality of the processing result of the substrate predicted by the result prediction unit, and a presentation unit that presents the processing conditions corresponding to the processing result evaluated as good by the evaluation unit.

[0127] In this case, the user can easily recognize the processing conditions when good processing results are obtained. As a result, the optimal processing conditions for the substrate can be easily explored.

[0128] (Item 13) The substrate processing system according to Item 13 includes the information processing apparatus according to any one of Items 10 to 12.

[0129] In this substrate processing system, sublimation drying of the substrate can be performed under optimal processing conditions based on the processing result of the substrate predicted by the information processing apparatus.

[0130] (Item 14) The learning method according to Item 14 obtaining data showing the relationship between the processing conditions in the process of sublimation drying a substrate using a processing liquid containing a sublimating agent and a solvent and at least one molecular descriptor of the sublimating agent and the solvent, and the processing result of the substrate; complementing the deficiency of at least one physical property value of the sublimating agent and the solvent in the obtained data using a prediction algorithm for predicting physical property values from molecular descriptors; generating a result model for predicting the processing result of the substrate by machine learning the data with the physical property values complemented, and is executed by a processor.

[0131] According to this learning method, a result model for predicting the processing result of the substrate can be generated by the processor performing machine learning on the data with the physical property values complemented. As a result, the optimal processing conditions can be explored based on the processing result of the substrate. Therefore, even when there is a deficiency in the physical property value of the sublimating agent or the solvent, it becomes possible to generate a result model for exploring the optimal processing conditions of the substrate.

[0132] (Item 15) The recipe determination method according to Item 15 is a recipe determination method using the result model generated by the learning method according to Item 14, obtaining the result model, determining a sublimating agent to be used for sublimation drying of a substrate; determining a solvent to be used for sublimation drying of a substrate; selecting processing conditions; predicting a processing result of a substrate from the determined sublimating agent, the determined solvent, and the selected processing conditions by using the obtained result model; and being executed by a processor.

[0133] According to this recipe determination method, based on the result model generated by the above learning method, the processing result of the substrate is predicted from the sublimating agent, the solvent, and the processing conditions used for sublimation drying of the substrate. Therefore, based on the predicted processing result of the substrate, the optimum processing conditions of the substrate can be searched for.

[0134] (Item 16) The learning program according to Item 16 a process of obtaining data indicating the relationship between the processing conditions in the process of sublimation drying a substrate using a processing liquid containing a sublimating agent and a solvent, at least one molecular descriptor of the sublimating agent and the solvent, and the processing result of the substrate; a process of complementing the lack of at least one physical property value of the sublimating agent and the solvent in the obtained data by using a prediction algorithm for predicting a physical property value from a molecular descriptor; a process of generating a result model for predicting the processing result of a substrate by machine learning the data with the physical property value complemented; and causing a processor to execute.

[0135] According to this learning program, a result model for predicting the processing result of a substrate can be generated by machine learning, by a processor, the data with the physical property value complemented. Thereby, based on the processing result of the substrate, the optimum processing conditions can be searched for. Therefore, even when there is a lack in the physical property value of the sublimating agent or the solvent, it becomes possible to generate a result model for searching for the optimum processing conditions of the substrate.

Explanation of Signs

[0136] 2... processing unit, 3... control device, 4... chamber, 10... spin chuck, 11... chuck pin, 12... spin base, 13... spin axis, 14... spin motor, 21... processing cup, 22... outer wall member, 23... cup, 24... guard, 27... guard lifting unit, 31... chemical solution nozzle, 34, 38, 42, 46... nozzle moving unit, 35... rinse solution nozzle, 39... processing solution nozzle, 40... processing solution pipe, 41... processing solution valve, 43... replacement solution nozzle, 51... blocking member, 51L... lower surface, 52... disc portion, 53... support shaft, 54... blocking member lifting unit, 55... center nozzle, 56... upper gas pipe, 57... upper gas valve, 58... flow rate adjustment valve, 59... upper temperature regulator, 100... substrate processing apparatus, 200... database storage device, 201... reference sublimating agent database, 202... reference solvent database, 203... result database, 204... processing sublimating agent database, 205... processing solvent database, 206... complementary sublimating agent database, 207... complementary solvent database, 208... integrated database, 209... learning database, 300... information processing apparatus, 301... model acquisition unit, 302... sublimating agent determination unit, 303... solvent determination unit, 304... condition selection unit, 305... result prediction unit, 306... evaluation unit, 307... presentation unit, 310... CPU, 320... RAM, 330... ROM, 340... storage device, 350... operation unit, 360... display device, 370... input / output I / F, 380... bus, 400... learning device, 401... reference sublimating agent acquisition unit, 402... sublimating agent model generation unit, 403... reference solvent acquisition unit, 404... solvent model generation unit, 405... result acquisition unit, 406... sublimating agent specifying unit, 407... solvent specifying unit, 408... sublimating agent complementary unit, 409... solvent complementary unit, 410... database generation unit, 411... processing solution calculation unit, 412... result model generation unit, 413... optimization unit, 500... substrate processing system, 501... network, A1... rotation axis, W... substrate

Claims

1. A data acquisition unit that acquires data showing the relationship between the processing conditions in the process of sublimation-drying a substrate using a processing liquid containing a sublimating agent and a solvent, at least one molecular descriptor of the sublimating agent and the solvent, and the processing result of the substrate; A data complementation unit that complements the lack of at least one physical property value of the sublimating agent and the solvent in the data acquired by the data acquisition unit using a prediction algorithm that predicts physical property values from molecular descriptors; A learning device comprising a result model generation unit that generates a result model for predicting the processing result of a substrate by machine-learning the data whose physical property values have been complemented by the data complementation unit.

2. The learning device according to claim 1, wherein the physical property value of the sublimating agent to be complemented by the data complementation unit includes at least one of vapor pressure, melting point, heat of dissolution, and Hansen solubility parameter.

3. The learning device according to claim 1 or 2, wherein the physical property value of the solvent to be complemented by the data complementation unit includes at least one of vapor pressure, Hansen solubility parameter, viscosity, and surface tension.

4. Further comprising a processing liquid calculation unit that calculates the molecular descriptor and physical property value of the processing liquid based on the molecular descriptor and physical property value of the sublimating agent and the molecular descriptor and physical property value of the solvent, The learning device according to claim 1 or 2, wherein the result model predicts the processing result of the substrate further based on the molecular descriptor and physical property value of the processing liquid calculated by the processing liquid calculation unit.

5. The learning device according to claim 4, wherein the physical property value of the processing liquid includes the solubility of the sublimating agent in the solvent.

6. The learning device according to claim 1 or 2, wherein the result model generation unit performs machine learning with at least a part of the processing conditions, the molecular descriptor and physical property value of the sublimating agent, and the molecular descriptor and physical property value of the solvent as explanatory variables and the processing result of the substrate as the objective variable.

7. Further comprising an optimization unit that optimizes the explanatory variables based on the importance of the explanatory variables, The learning device according to claim 6, wherein the result model generation unit performs machine learning on the explanatory variables optimized by the optimization unit.

8. Further comprising a sublimating agent model generation unit that generates a sublimating agent model for predicting physical property values by machine-learning the relationship between the molecular descriptor and physical property value of a reference sublimating agent, The data complementation unit includes a sublimating agent complementation unit that predicts physical property values from the molecular descriptors of the sublimating agent using the sublimating agent model generated by the sublimating agent model generation unit, and complements the missing physical property values of the sublimating agent in the data acquired by the data acquisition unit using the predicted physical property values. The learning device according to claim 1 or 2.

9. Further comprising a solvent model generation unit that generates a solvent model for predicting physical property values by machine learning the relationship between the molecular descriptors of the reference solvent and the physical property values, The data complementation unit includes a solvent complementation unit that predicts physical property values from the molecular descriptors of the solvent using the solvent model generated by the solvent model generation unit, and complements the missing physical property values of the solvent in the data acquired by the data acquisition unit using the predicted physical property values. The learning device according to claim 1 or 2.

10. An information processing device using the result model generated by the learning device according to claim 1 or 2, A model acquisition unit that acquires the result model, A sublimating agent determination unit that determines a sublimating agent used for sublimation drying of the substrate, A solvent determination unit that determines a solvent used for sublimation drying of the substrate, A condition selection unit that selects processing conditions, A result prediction unit that predicts the processing result of the substrate from the sublimating agent determined by the sublimating agent determination unit, the solvent determined by the solvent determination unit, and the processing conditions selected by the condition selection unit using the result model acquired by the model acquisition unit. An information processing device.

11. The condition selection unit selects processing conditions using Bayesian optimization, the steepest descent method, or a genetic algorithm. The information processing device according to claim 10.

12. An evaluation unit that evaluates the quality of the processing result of the substrate predicted by the result prediction unit, The information processing device according to claim 10, further comprising a presentation unit that presents the processing conditions corresponding to the processing results evaluated as good by the evaluation unit.

13. A substrate processing system comprising the information processing device according to claim 10.

14. Obtaining data showing the relationship between the processing conditions in the process of sublimation drying the substrate using a processing liquid containing a sublimating agent and a solvent, at least one molecular descriptor of the sublimating agent and the solvent, and the processing result of the substrate; Complementing the missing physical property values of at least one of the sublimating agent and the solvent in the acquired data using a prediction algorithm for predicting physical property values from molecular descriptors. Generating a result model for predicting the processing result of a substrate by machine learning data with physical property values complemented. A learning method executed by a processor.

15. A recipe determination method using the result model generated by the learning method according to Claim 14, comprising: Obtaining the result model; Determining a sublimating agent used for sublimation drying of the substrate; Determining a solvent used for sublimation drying of the substrate; Selecting processing conditions; Predicting the processing result of the substrate from the determined sublimating agent, the determined solvent, and the selected processing conditions using the obtained result model. A recipe determination method executed by a processor.

16. A process of obtaining data showing the relationship between the processing conditions in the process of sublimation drying a substrate using a processing liquid containing a sublimating agent and a solvent, at least one molecular descriptor of the sublimating agent and the solvent, and the processing result of the substrate; A process of complementing the lack of physical property values of at least one of the sublimating agent and the solvent in the obtained data using a prediction algorithm for predicting physical property values from molecular descriptors; A process of generating a result model for predicting the processing result of a substrate by machine learning using the data with physical property values complemented. A learning program to be executed by a processor.

Images