Substrate processing apparatus control method and substrate processing system

Abstract

The invention provides a control method of a substrate processing apparatus and a substrate processing system. Comprising the following steps of: a) locally etching a processed body to form a concave part in the processed body; b) forming a protective film on a side wall of a recess formed in the object to be processed; c) further etching the object to be processed on which the recess and the protective film are formed; d) repeating said b) and said c); e) monitoring the object to be processed obtained by at least one of the steps a) to d); f) performing a virtual experiment for simulating the steps a) to d); g) deriving a parameter to be applied to at least one of the a) to the e) based on a monitoring result of the object to be processed and a result of the virtual experiment; and h) implementing at least one of said a) to said d) to which the derived parameter is applied.

Description

Technical Field

[0001] The present disclosure relates to a control method of a substrate processing device and a substrate processing system. Background Art

[0002] As semiconductor devices are integrated not only in the horizontal direction but also in the vertical direction, the aspect ratio of the pattern formed in the manufacturing process of the semiconductor device is also increasing. For example, in the manufacture of 3D NAND, a channel via is formed in a direction that penetrates a large number of metal wiring layers. If a 64-layer memory cell is formed, the aspect ratio of the channel via will be 45.

[0003] Various methods have been proposed to form a pattern with a high aspect ratio with high precision. For example, a method has been proposed to suppress lateral etching by repeatedly performing etching and film formation on an opening of a dielectric material formed in a substrate. In addition, a method has been proposed to form a protective film for preventing lateral etching of a dielectric layer by combining etching and film formation (for example, refer to Patent Document 1).

[0004] Prior art literature

[0005] Patent Literature

[0006] Patent Document 1: U.S. Patent Application Publication No. 2018 / 0247798 Summary of the invention

[0007] Problem that the invention aims to solve

[0008] The present disclosure provides a control method of a substrate processing device and a substrate processing system capable of efficiently guiding a substrate to have a good opening shape.

[0009] Solutions for solving problems

[0010] A control method for a substrate processing device involved in one embodiment of the present invention includes the following steps: a) locally etching a processed body to form a recess in the processed body; b) forming a protective film on the side wall of the recess formed in the processed body; c) further etching the processed body formed with the recess and the protective film; d) repeating b) and c); e) monitoring the processed body obtained by at least one of a) to d); f) performing a virtual experiment simulating a virtual experiment of a) to d); g) deriving parameters that should be applied to at least one of a) to d) based on the monitoring results of the processed body and the results of the virtual experiment; and h) implementing at least one of a) to d) that applies the derived parameters.

[0011] Effects of the Invention

[0012] According to the present disclosure, a good opening shape can be efficiently guided. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a diagram for explaining a configuration example of a substrate processing system.

[0014] Figure 2 It is a diagram for explaining a configuration example of a capacitive coupling type plasma processing apparatus.

[0015] Figure 3 This is a flowchart illustrating a general ALD process.

[0016] Figure 4 1 is an explanatory diagram for explaining a first method of subconformal ALD.

[0017] Figure 5 1 and 2 are diagrams for explaining a second method of subconformal ALD.

[0018] Figure 6 This is a flowchart illustrating the flow of ALD in this embodiment.

[0019] Figure 7 This is a flowchart illustrating the flow of ALD in this embodiment.

[0020] Figure 8 It is an explanatory diagram for explaining a configuration example of an etching simulator.

[0021] Fig. 9 is a flow chart illustrating the updating process of the model.

[0022] Fig.10 It is an explanatory diagram for explaining the film thickness of the protective film formed by the substrate processing method according to the embodiment.

[0023] Fig.11 This is a schematic diagram of the reinforcement learning algorithm.

[0024] Fig.12 1 is a schematic diagram showing a first structural example of a plasma processing system in a third embodiment.

[0025] Fig.13 This is a schematic diagram showing a second configuration example of the plasma processing system in the third embodiment. DETAILED DESCRIPTION

[0026] Hereinafter, the present invention will be described in detail based on the drawings showing embodiments of the present invention.

[0027] (Implementation Method 1)

[0028] Figure 11 is a diagram for explaining a structural example of a substrate processing system. The substrate processing system PS includes stations BA1 to BA1, containers RC1 to RC4, a loader module LM, an aligner AN, load lock modules LL1 and LL2, process modules PM1 to PM6, a transfer module TF, and a control device MC. In addition, the number of stations, the number of containers, and the number of load lock modules in the substrate processing system PS can be any number of more than one. In addition, the number of process modules in the substrate processing system PS can be any number of more than two.

[0029] The tables BA1 to BA4 are arranged along one edge of the loader module LM. The containers RC1 to RC4 are respectively mounted on the tables BA1 to BA4. The containers RC1 to RC4 are, for example, containers called FOUPs (Front Opening Unified Pods). The containers RC1 to RC4 are respectively configured to accommodate substrates W therein.

[0030] The loader module LM has a chamber. The pressure in the chamber of the loader module LM is set to atmospheric pressure. The loader module LM has a transport device TU1. The transport device TU1 is, for example, a multi-joint robot, and is controlled by the control device MC. The transport device TU1 is configured to transport the substrate W through the chamber of the loader module LM. The transport device TU1 can transport the substrate W between each container RC1 to RC4 and the aligner AN, between the aligner AN and each load interlock module LL1 to LL2, and between each load interlock module LL1 to LL2 and each container RC1 to RC4. The aligner AN is connected to the loader module LM. The aligner AN is configured to adjust the position of the substrate W (calibrate the position).

[0031] The load lock module LL1 and the load lock module LL2 are respectively provided between the loader module LM and the transfer module TF. The load lock module LL1 and the load lock module LL2 respectively provide a pre-decompression chamber.

[0032] The transport module TF is connected to the load interlock module LL1 and the load interlock module LL2 respectively via gate valves. The transport module TF has a transport chamber TC capable of reducing pressure. The transport module TF has a transport device TU2. The transport device TU2 is, for example, a multi-joint robot, which is controlled by a control device MC. The transport device TU2 is configured to transport the substrate W via the transport chamber TC. The transport device TU2 can transport the substrate W between each load interlock module LL1~LL2 and each process module PM1~PM6, and between any two process modules among the process modules PM1~PM6.

[0033] Each process module PM1 to PM6 is a processing device configured to perform dedicated substrate processing. One of the process modules PM1 to PM6 is a film forming device. The film forming device is used to form a protective film PF in the film forming process described later. When plasma is generated in the film forming process, the film forming device is a plasma processing device having a structure for generating plasma. When the protective film PF is formed in a manner that does not generate plasma in the film forming process, the film forming device may not have a structure for generating plasma. The other process modules among the process modules PM1 to PM6 are etching devices. The etching device is used to form a pattern on the surface of the object to be processed in the etching process described later.

[0034] In the substrate processing system PS, the control device MC is configured to control each part of the substrate processing system PS. For example, the control device MC can control the operation of the etching device to form a recess in the processed object to form a pattern on the surface of the processed object. In addition, the control device MC can control the film forming device to form a protective film on the side wall of the formed recess.

[0035] The substrate processing system PS is equipped with an observation device OC. The observation device OC can be set at any place in the substrate processing system PS. In one example, the observation device OC is set in the observation module OM adjacent to the loader module LM. The substrate W can be moved between the observation module OM and the process modules PM1 to PM6 by the conveying device TU1 and the conveying device TU2. After the substrate W is accommodated in the observation module OM by the conveying device TU1 and the position of the substrate W is aligned in the observation module OM, the observation device OC measures the groove width of the pattern of the mask, etc. of the substrate W, and sends the measurement result to the control device MC. In the observation device OC, the groove width of the pattern of the mask, etc. formed in multiple areas on the surface of the substrate W can be measured. As the observation device OC, for example, an optical observation device, a weight meter, an ultrasonic microscope, etc. can be used.

[0036] Next, a configuration example of a plasma processing system will be described. Figure 2 It is a diagram for explaining a configuration example of a capacitive coupling type plasma processing apparatus.

[0037] The plasma processing system includes a capacitively coupled plasma processing device 1 and a control unit 2. The capacitively coupled plasma processing device 1 includes a plasma processing chamber 10, a gas supply unit 20, a power supply 30, and an exhaust system 40. In addition, the plasma processing device 1 includes a substrate support unit 11 and a gas introduction unit. The gas introduction unit is configured to introduce at least one processing gas into the plasma processing chamber 10. The gas introduction unit includes a shower head 13. The substrate support unit 11 is disposed in the plasma processing chamber 10. The shower head 13 is disposed above the substrate support unit 11. In one embodiment, the shower head 13 constitutes at least a portion of the ceiling of the plasma processing chamber 10. The plasma processing chamber 10 has a plasma processing space 10s defined by the shower head 13, the side wall 10a of the plasma processing chamber 10, and the substrate support unit 11. The plasma processing chamber 10 has at least one gas supply port for supplying at least one processing gas to the plasma processing space 10s, and at least one gas exhaust port for exhausting gas from the plasma processing space. The plasma processing chamber 10 is grounded. The shower head 13 and the substrate support 11 are electrically insulated from the housing of the plasma processing chamber 10 .

[0038] The substrate support portion 11 includes a main body 111 and a ring assembly 112. The main body 111 has a central area 111a for supporting a substrate W and an annular area 111b for supporting the ring assembly 112. A wafer is an example of a substrate W. The annular area 111b of the main body 111 surrounds the central area 111a of the main body 111 when viewed from above. The substrate W is arranged on the central area 111a of the main body 111, and the ring assembly 112 is arranged on the annular area 111b of the main body 111 in a manner that surrounds the substrate W on the central area 111a of the main body 111. Therefore, the central area 111a is also referred to as a substrate supporting surface for supporting the substrate W, and the annular area 111b is also referred to as a ring supporting surface for supporting the ring assembly 112.

[0039] In one embodiment, the main body 111 includes a base 1110 and an electrostatic holding disk 1111. The base 1110 includes a conductive component. The conductive component of the base 1110 can function as a lower electrode. The electrostatic holding disk 1111 is arranged on the base 1110. The electrostatic holding disk 1111 includes a ceramic component 1111a and an electrostatic electrode 1111b arranged in the ceramic component 1111a. The ceramic component 1111a has a central area 111a. In one embodiment, the ceramic component 1111a also has an annular area 111b. In addition, other components surrounding the electrostatic holding disk 1111, such as an annular electrostatic holding disk or an annular insulating component, may also have an annular area 111b. In this case, the ring assembly 112 may be arranged on the annular electrostatic holding disk or the annular insulating component, or on both the electrostatic holding disk 1111 and the annular insulating component. In addition, at least one RF / DC electrode coupled to the RF (Radio Frequency) power supply 31 and / or the DC (Direct Current) power supply 32 described later may also be arranged in the ceramic component 1111a. In this case, at least one RF / DC electrode functions as a lower electrode. When a bias RF signal and / or a DC signal described later is supplied to at least one RF / DC electrode, the RF / DC electrode is also referred to as a bias electrode. In addition, the conductive component of the base 1110 and at least one RF / DC electrode may also function as a plurality of lower electrodes. In addition, the electrostatic electrode 1111b may also function as a lower electrode. Therefore, the substrate support portion 11 includes at least one lower electrode.

[0040] The ring assembly 112 includes one or more ring-shaped members. In one embodiment, the one or more ring-shaped members include one or more edge rings and at least one cover ring. The edge ring is formed of a conductive material or an insulating material, and the cover ring is formed of an insulating material.

[0041] In addition, the substrate support portion 11 may also include a temperature control module configured to adjust at least one of the electrostatic holding disk 1111, the ring assembly 112, and the substrate to a target temperature. The temperature control module may include a heater, a heat transfer medium, a flow path 1110a, or a combination thereof. A heat transfer fluid such as brine or gas flows in the flow path 1110a. In one embodiment, the flow path 1110a is formed in the base 1110, and one or more heaters are arranged in the ceramic component 1111a of the electrostatic holding disk 1111. In addition, the substrate support portion 11 may also include a heat transfer gas supply portion configured to supply a heat transfer gas to the gap between the back side of the substrate W and the central area 111a.

[0042] The shower head 13 is configured to introduce at least one processing gas from the gas supply unit 20 into the plasma processing space 10s. The shower head 13 has at least one gas supply port 13a, at least one gas diffusion chamber 13b, and a plurality of gas introduction ports 13c. The processing gas supplied to the gas supply port 13a is introduced into the plasma processing space 10s from the plurality of gas introduction ports 13c through the gas diffusion chamber 13b. In addition, the shower head 13 includes at least one upper electrode. In addition, the gas introduction unit may also include, in addition to the shower head 13, one or more side gas injection units (SGI: Side Gas Injector) installed in one or more openings formed in the side wall 10a.

[0043] The gas supply unit 20 may include at least one gas source 21 and at least one flow controller 22. In one embodiment, the gas supply unit 20 is configured to supply at least one processing gas from the respective corresponding gas sources 21 to the shower head 13 via the respective corresponding flow controllers 22. Each flow controller 22 may include, for example, a mass flow controller or a pressure-controlled flow controller. In addition, the gas supply unit 20 may also include one or more flow modulation devices for modulating or pulsing the flow of at least one processing gas.

[0044] The power supply 30 includes an RF power supply 31 coupled to the plasma processing chamber 10 via at least one impedance matching circuit. The RF power supply 31 is configured to supply at least one RF signal (RF power) to at least one lower electrode and / or at least one upper electrode. Thus, plasma is formed from at least one processing gas supplied to the plasma processing space 10s. Therefore, the RF power supply 31 can function as at least a part of a plasma generating unit configured to generate plasma from one or more processing gases in the plasma processing chamber 10. In addition, by supplying a bias RF signal to at least one lower electrode, a bias potential can be generated on the substrate W, thereby attracting ion components in the formed plasma to the substrate W.

[0045] In one embodiment, the RF power supply 31 includes a first RF generating unit 31a and a second RF generating unit 31b. The first RF generating unit 31a is coupled to at least one lower electrode and / or at least one upper electrode via at least one impedance matching circuit, and is configured to generate a source RF signal (source RF power) for plasma generation. In one embodiment, the source RF signal has a frequency in the range of 10 MHz to 150 MHz. In one embodiment, the first RF generating unit 31a may also be configured to generate a plurality of source RF signals having different frequencies. The generated one or more source RF signals are supplied to at least one lower electrode and / or at least one upper electrode.

[0046] The second RF generating unit 31b is coupled to at least one lower electrode via at least one impedance matching circuit, and is configured to generate a bias RF signal (bias RF power). The frequency of the bias RF signal may be the same as the frequency of the source RF signal, or may be different from the frequency of the source RF signal. In one embodiment, the bias RF signal has a frequency lower than the frequency of the source RF signal. In one embodiment, the bias RF signal has a frequency in the range of 100kHz to 60MHz. In one embodiment, the second RF generating unit 31b may also be configured to generate a plurality of bias RF signals having different frequencies. The generated one or more bias RF signals are supplied to at least one lower electrode. In addition, in various embodiments, at least one of the source RF signal and the bias RF signal may also be pulsed.

[0047] In addition, the power supply 30 may also include a DC power supply 32 coupled to the plasma processing chamber 10. The DC power supply 32 includes a first DC generating unit 32a and a second DC generating unit 32b. In one embodiment, the first DC generating unit 32a is connected to at least one lower electrode and is configured to generate a first DC signal. The generated first DC signal is applied to at least one lower electrode. In one embodiment, the second DC generating unit 32b is connected to at least one upper electrode and is configured to generate a second DC signal. The generated second DC signal is applied to at least one upper electrode.

[0048] In various embodiments, at least one of the first DC signal and the second DC signal may also be pulsed. In this case, a sequence of pulse voltages is applied to at least one lower electrode and / or at least one upper electrode. The pulse voltage may have a pulse waveform of a rectangle, a trapezoid, a triangle, or a combination thereof. In one embodiment, a waveform generator for generating a sequence of pulse voltages from a DC signal is connected between the first DC generator 32a and at least one lower electrode. Thus, the first DC generator 32a and the waveform generator constitute a pulse voltage generator. In the case where the pulse voltage generator is constituted by the second DC generator 32b and the waveform generator, the pulse voltage generator is connected to at least one upper electrode. The pulse voltage may have both a positive polarity and a negative polarity. In addition, the sequence of pulse voltages may also include one or more positive polarity pulse voltages and one or more negative polarity pulse voltages in one cycle. In addition, a first DC generator 32a and a second generator 32b may be provided in addition to the RF power supply 31, or a first DC generator 32a may be provided to replace the second RF generator 31b.

[0049] The exhaust system 40 can be connected to, for example, a gas exhaust port 10e disposed at the bottom of the plasma processing chamber 10. The exhaust system 40 can include a pressure regulating valve and a vacuum pump. The pressure in the plasma processing space 10s is adjusted by the pressure regulating valve. The vacuum pump can include a turbomolecular pump, a dry pump, or a combination thereof.

[0050] The plasma processing device 1 is provided with a light sensor 108 capable of measuring the intensity of light of each wavelength in the plasma in the plasma processing space 10s through the quartz window 109. The light sensor 108 includes a first sensor 108a and a second sensor 108b. The first sensor 108a is a sensor for sensing the state of the plasma generated in the plasma processing space 10s. On the other hand, the second sensor 108b is a sensor for sensing the pattern shape on the surface of the substrate W placed on the base 1110. The sensing data of the first sensor 108a and the second sensor 108b are output to the control unit 2. The control unit 2 measures / estimates the plasma state in the plasma processing chamber 10 and the pattern shape of the substrate W based on the sensing data of the first sensor 108a and the second sensor 108b.

[0051] The control unit 2 processes computer-executable commands for causing the plasma processing device 1 to perform various processes described in the present disclosure. The control unit 2 can be configured to control the various elements of the plasma processing device 1 to perform various processes described herein. In one embodiment, a part or all of the control unit 2 may also be included in the plasma processing device 1. The control unit 2 may include a processing unit 2a1, a storage unit 2a2, and a communication interface 2a3. The control unit 2 is implemented, for example, by a computer 2a. The processing unit 2a1 can be configured to perform various control actions by reading a program from the storage unit 2a2 and executing the read program. The program can be stored in the storage unit 2a2 in advance, or it can be obtained via a medium when necessary. The obtained program is stored in the storage unit 2a2, and is read and executed from the storage unit 2a2 by the processing unit 2a1. The medium can be various storage media that can be read by the computer 2a, or it can be a communication line connected to the communication interface 2a3. The processing unit 2a1 can be a CPU (Central Processing Unit). The storage unit 2a2 may include a RAM (Random Access Memory), a ROM (Read Only Memory), a HDD (Hard Disk Drive), an SSD (Solid State Drive), or a combination thereof. The communication interface 2a3 may communicate with the plasma processing apparatus 1 via a communication line such as a LAN.

[0052] The programs stored in the storage unit 2a2 include simulators for virtual experiments for simulating processes (actual experiments) performed in the substrate processing system PS. Simulators for virtual experiments include plasma simulators, reaction product simulators, shape simulators, etc. In addition, the programs stored in the storage unit 2a2 may also be programs for implementing virtual metrology (VM: Virtual Metrology) technology. These computer programs may be a single computer program or a program composed of multiple computer programs. In addition, these computer programs may also be programs that partially use existing program libraries.

[0053] Next, the process implemented in the substrate processing system PS is described. When a pattern such as HARC (High Aspect Ratio Contact) is formed on a substrate as a processed object, a shape abnormality called bowing sometimes occurs. Bowing refers to the phenomenon that when an opening is formed in the longitudinal direction (the thickness direction of the substrate), the inner peripheral surface of the opening also bulges in the lateral direction (the in-plane direction of the substrate). In order to suppress the occurrence of shape abnormalities, a method of forming a protective film on the side wall of the opening is proposed. As a film forming method for such a protective film, for example, atomic layer deposition (Atomic Layer Deposition: ALD) is used. In this embodiment, the film forming process using ALD is mainly described, but as an alternative, other film forming methods such as plasma-enhanced ALD (Plasma-enhanced ALD: PEALD), chemical vapor deposition (Chemical Vapor Deposition: CVD), plasma-enhanced CVD (Plasma-enhanced CVD: PECVD), and plasma annular chemical vapor deposition (PECCVD) can be used.

[0054] In addition, in the following description, "pattern" refers to all shapes formed on the substrate. A pattern refers to a plurality of shapes formed on the substrate as a whole, such as holes, grooves, line widths and line spacings. In addition, a "recess" refers to a shape portion formed in a pattern on the substrate that is recessed in the thickness direction of the substrate. In addition, the recess has a "side wall" which is the inner peripheral surface of the recessed shape, a "bottom" which is the bottom portion of the recessed shape, and a "top" which is continuous with the side wall and is the surface of the substrate near the side wall. In addition, the space surrounded by the top is referred to as an "opening". In addition, the term "opening" is also used to refer to the entire space surrounded by the bottom and side walls of the recess or any position of the space.

[0055] Figure 31 is a flowchart for explaining a general ALD process. The substrate processing system PS provides a processing object to the interior of the plasma processing chamber 10 (step S11). The processing object is, for example, a substrate on which a pattern with a high aspect ratio is formed by other process modules. Alternatively, the processing object may be a substrate without a pattern formed thereon. In this case, in the plasma processing apparatus 1, the processing object may be partially etched to form a pattern.

[0056] The plasma processing apparatus 1 introduces a first gas into the plasma processing chamber 10 (step S12). The first gas is also called a precursor. The plasma processing apparatus 1 purges the plasma processing chamber 10 to discharge components of the first gas excessively adsorbed on the surface of the object to be processed (step S13).

[0057] The plasma processing apparatus 1 introduces a second gas into the plasma processing chamber 10 to generate plasma of the second gas (step S14). The second gas is also called a reaction gas. The plasma processing apparatus 1 purges the plasma processing chamber 10 to discharge excess components of the second gas (step S15).

[0058] Through the processes of steps S12 to S15, a protective film is formed on the side wall of the opening. After the protective film is formed, the plasma processing apparatus 1 etches the object to be processed (step S16).

[0059] In addition, Figure 3 In the flowchart, it is assumed that both the film forming process and the etching process are performed in one plasma processing apparatus 1, but the film forming process may be performed in one plasma processing apparatus and the etching process may be performed in another plasma processing apparatus. In addition, the film forming process and the etching process may be performed in the same substrate processing system or in different substrate processing systems.

[0060] In addition, the plasma processing device 1 may also measure the film thickness of the protective film after step S15 to determine whether the desired film thickness is obtained. If the desired film thickness is not obtained, the plasma processing device 1 returns the process to step S12 to continue forming the protective film. For measuring the film thickness, the second sensor 108b is used in the case of in-situ, and the observation device OC is used in the case of ex-situ.

[0061] In addition, the plasma processing device 1 may also measure the shape of the pattern after step S16 to determine whether the desired shape is obtained. If the desired shape is not obtained, the plasma processing device 1 returns the process to step S12 to continue the formation and etching of the protective film. For the measurement of the shape, the second sensor 108b is used in the case of in-situ, and the observation device OC is used in the case of non-in-situ.

[0062] In ALD, a film is formed by self-limiting adsorption and reaction of a specific component to a substance existing on the substrate surface. Therefore, in ALD, conformal film formation can be achieved by setting a sufficient processing time. For example, Figure 3 In the flowchart of , when the processing time of steps S12 and S14 is sufficiently long (when the processing conditions are set to saturation conditions), the adsorption of the components of the first gas to the substrate and the reaction of the components of the first gas with the components of the second gas reach saturation, and a conformal film is formed. A conformal film refers to a film having the same thickness regardless of any position on the substrate (for example, a position in the up and down directions).

[0063] In contrast, in subconformal ALD, while using the same process as ALD, the process conditions are controlled so that at least one of the adsorption and reaction of the film-forming components does not reach saturation. That is, in subconformal ALD, a subconformal film is formed by not completing the self-limiting adsorption or reaction on the surface of the substrate. A subconformal film is a film whose film thickness changes depending on the position on the substrate (for example, the position in the up and down direction). In one example, it may be a film that is thick on the upper side (opening side) and thin on the lower side, or it may be a film whose film thickness decreases as it moves from the upper side to the lower side.

[0064] There are at least two methods for sub-conformal ALD processing.

[0065] (1) The precursor is adsorbed on the entire surface of the substrate. Control is performed so that the reaction gas introduced later does not spread over the entire surface of the wafer.

[0066] (2) The precursor is adsorbed only on a portion of the surface of the substrate. The reaction gas introduced thereafter forms a film only on the surface portion where the precursor is adsorbed.

[0067] By using the method (1) or (2) described above, a film whose thickness gradually decreases from top to bottom can be formed on the side wall of a pattern formed on a substrate.

[0068] Figure 4 1 is an explanatory diagram for explaining a first method of subconformal ALD. Figure 4 The object to be processed shown in the figure includes an etching target film EL1 and a mask MA. A recessed portion having an opening OP is formed in a stacked body of the etching target film EL1 and the mask MA.

[0069] The plasma processing apparatus 1 introduces a precursor P into a plasma processing chamber 10 in which a target object is disposed. Figure 4 In the first method, a sufficient treatment time is set for the adsorption of the precursor P. Thus, the precursor P is adsorbed on the entire surface of the object to be processed ( Figure 4 After the adsorption of the precursor P is completed, the plasma processing apparatus 1 purges the plasma processing chamber 10, and then introduces the reaction gas R into the plasma processing chamber 10 ( Figure 4 (C)). The introduced reaction gas R reacts with the precursor P on the object to be processed to gradually form a protective film PF from above the mask MA. Here, before the formation of the protective film PF reaches below the etching target film EL1, the reaction gas R is purged. By such a process, the protective film PF is not formed on the entire side wall of the recessed portion while using the ALD method, but is formed only on the mask MA and the upper portion of the etching target film EL1 ( Figure 4 (D)). Figure 4 In (D), the protection film PF is formed above the side wall and the top of the recessed portion, but is not formed below the side wall and the bottom.

[0070] Figure 5 1 and 2 are diagrams for explaining a second method of subconformal ALD. Figure 5 The shape of the object to be processed is shown in FIG. Figure 4 The shapes of the treated objects shown are the same.

[0071] In the second method, the plasma processing apparatus 1 allows the precursor P to be adsorbed only on the upper portion of the object to be processed ( Figure 5 After the precursor P is purged, the plasma processing apparatus 1 introduces a reaction gas R into the plasma processing chamber 10 ( Figure 5 At this time, the reaction gas R reacts only at the position where the precursor P is adsorbed to form a film, so the protective film PF is formed only above the object to be processed ( Figure 5 (C)).

[0072] Figure 4 Eligible for execution under unsaturated conditions Figure 3 For example, in step S14, Figure 5 Eligible for execution under unsaturated conditions Figure 3 When the processing time of step S12 and step S14 is sufficiently long, the formed film becomes conformal. Therefore, in subconformal ALD, the processing conditions are set so that at least one of the adsorption and reaction of the film-forming components does not reach saturation.

[0073] The processing conditions adjusted to achieve subconformal ALD include, for example, the temperature of the substrate support 11 on which the substrate W is placed, the pressure in the plasma processing chamber 10, the flow rate and introduction time of the introduced precursor, the gas flow rate and introduction time of the introduced reaction gas, the processing time, etc. In addition, in the case of a process using plasma, the film formation position can also be adjusted by adjusting the value of the high frequency (RF) power applied to generate the plasma. Figure 3 In the case of the process of, the second gas is converted into plasma in step S14, but the first gas in step S12 may also be converted into plasma.

[0074] Figure 6 and Figure 7 : is a flowchart illustrating the process of ALD in this embodiment. The substrate processing system PS provides a processing object to the interior of the plasma processing chamber 10 (step S101). The processing object is a substrate on which no pattern is formed. In the plasma processing device 1, sensing by the second sensor 108b is performed at any time, and the sensing data obtained by the second sensor 108b is output to the control unit 2. The control unit 2 measures / estimates the shape of the surface of the processing object provided to the interior of the plasma processing chamber 10 based on the sensor data of the second sensor 108b (step S102). The shape measured / estimated in step S102 may be the shape of each concave portion formed on the surface of the processing object, or may be the uniformity of the overall concave shape in the surface of the processing object.

[0075] In addition to the sensor data of the second sensor 108b, the control unit 2 also inputs information related to the substrate to be processed, the processing conditions of the etching process, various output data output from the plasma processing device 1, and various measurement data measured during the execution of the etching process. Here, the information related to the substrate to be processed includes information such as the material, thickness, aspect ratio, and mask coverage of the substrate. The processing conditions of the etching process include information such as the pressure in the chamber, the power of the high-frequency power supply, the gas flow rate, the gas mixing ratio, the temperature in the chamber, and the temperature of the surface of the processed object. The output data of the plasma processing device 1 includes data such as the source RF power, the bias RF power, and the luminous intensity obtained by the OES (Optical Emission Spectrometer). The measurement data during the execution of the process includes data such as the plasma density, ion energy, and ion flow rate.

[0076] The control unit 2 performs virtual etching (step S103) simulating etching in the plasma processing apparatus 1. The control unit 2 estimates the shape of the object after etching by simulation using the pattern shape measured / estimated in step S102 as the initial shape. The configuration example of the etching simulator will be described in detail later.

[0077] The control unit 2 obtains various parameters used in the virtual etching, and derives the parameters that should be applied to the actual experiment based on the obtained parameters (step S104). Here, the parameters used in the virtual etching include parameters of the substrate such as the material, thickness, aspect ratio, and mask coverage of the substrate, as well as parameters such as the pressure in the chamber, the power of the high-frequency power supply, the gas flow rate, the gas mixing ratio, the temperature in the chamber, the temperature of the surface of the processed object, the source RF power, the bias RF power, OES, the plasma density, the ion energy, and the ion flow rate. In addition, the types of parameters that should be applied to the actual experiment can be set in advance or selected by the control unit 2. For example, the control unit 2 compares the parameters set according to the actual experiment with the parameters obtained from the virtual experiment, and selects the parameters that should be applied to the actual experiment based on the difference between the two.

[0078] When the control unit 2 derives the parameters to be applied to the actual experiment in step S104, a learning model of machine learning including deep learning, reinforcement learning, etc., a statistical model, and a model based on a combination thereof, etc., can be used. These models are generated by finding a quantitative relationship between the parameters used in the virtual etching and the parameters to be applied to the actual experiment using a known method known in machine learning, statistical analysis, etc. The control unit 2 can derive the parameters to be applied to the actual experiment by inputting the parameters acquired in step S104 into the generated model.

[0079] Furthermore, the control unit 2 may optimize the parameters to be applied to the actual experiment so that the consistency between the shape measured / estimated in the actual experiment and the shape predicted in the virtual experiment becomes higher or the process time (throughput) is shortened.

[0080] The plasma processing apparatus 1 acquires the parameters derived from the virtual experiment in step S103, and performs etching using the acquired parameters (step S105). This etching process is a process in the actual experiment.

[0081] The control unit 2 acquires the sensing data output from the second sensor 108b during the etching (actual experiment). The control unit 2 measures / estimates the pattern shape of the pattern formed by the etching in step S105 based on the sensing data of the second sensor 108b (step S106). The shape measured / estimated in step S106 may be the shape of each concave portion formed on the surface of the processed object, or the uniformity of the overall concave shape in the surface of the processed object.

[0082] The control unit 2 determines whether the ideal shape is obtained based on the measurement / estimation result of the pattern shape (step S107). The control unit 2 determines whether the ideal shape is obtained by measuring / estimating the shape of the concave portion formed by etching based on the sensor data obtained from the second sensor 108b, and determining whether the concave portion has a desired aspect ratio. Alternatively, the control unit 2 may determine whether the ideal shape is obtained by measuring / estimating the opening width and opening depth of the concave portion formed by etching, and determining whether the opening width and opening depth are within a set range.

[0083] In addition, the control unit 2 may also compare the measurement / estimation result of the pattern shape with the set value set for the pattern shape, and stop the subsequent processing according to the comparison result. The set value is a value set for the aspect ratio, opening width, opening depth, etc. of the pattern shape. As an alternative, the control unit 2 may also output an alarm according to the comparison result of the measurement / estimation result of the pattern shape with the set value. For example, the control unit 2 outputs an alarm by notifying the terminal carried by the user via the communication interface 2a3 that the measurement / estimation result of the pattern shape exceeds the set value (or is less than the set value). In the case where the computer 2a has a display and a sound output unit, an alarm may also be output by displaying text information on the display or outputting sound from the sound output unit.

[0084] If it is determined that the ideal shape has not been obtained (S107: No), the control unit 2 returns the process to step S105 and continues etching the object to be processed until the ideal shape is obtained. At this time, the control unit 2 can also obtain various output data and various measurement data output from the plasma processing device 1 during the execution of etching at any time, and repeatedly perform virtual etching with reference to the obtained data. The control unit 2 can derive parameters that should be applied to the actual experiment from the virtual etching, and apply the parameters to the etching (actual experiment) repeatedly performed in the plasma processing device 1.

[0085] After executing the virtual etching in step S103, the control unit 2 executes the virtual ALD (step S108). The control unit 2 uses the pattern shape obtained by the virtual etching in step S103 as the initial shape to estimate the shape of the processed object after ALD through simulation. The configuration example of the ALD simulator will be described in detail later.

[0086] The control unit 2 obtains various parameters used in the virtual ALD, and derives parameters that should be applied to the actual experiment based on the obtained parameters (step S109). The parameters used in the virtual ALD are the same as those of the virtual etching, including the material, thickness, aspect ratio, mask coverage, pressure in the chamber, power of the high-frequency power supply, gas flow rate, gas mixing ratio, temperature in the chamber, temperature of the surface of the processed object, source RF power, bias RF power, OES, plasma density, ion energy, ion flow rate and other parameters of the substrate. The types of parameters that should be applied to the actual experiment can be set in advance or selected by the control unit 2. For example, the control unit 2 compares the parameters set according to the actual experiment with the parameters obtained from the virtual experiment, and selects the parameters that should be applied to the actual experiment based on the difference between the two.

[0087] When the control unit 2 derives the parameters to be applied to the actual experiment in step S109, a learning model including machine learning such as deep learning and reinforcement learning, a statistical model, and a model based on a combination thereof can be used. These models are generated by finding a quantitative relationship between the parameters used in the virtual ALD and the parameters to be applied to the actual experiment using a known method known in machine learning, statistical analysis, etc. The control unit 2 can derive the parameters to be applied to the actual experiment by inputting the parameters acquired in step S109 into the generated model.

[0088] Furthermore, the control unit 2 may optimize the parameters to be applied to the actual experiment so that the consistency between the shape measured / estimated in the actual experiment and the shape predicted in the virtual experiment becomes higher or the process time (throughput) is shortened.

[0089] When the ideal shape is obtained in step S107 (S107: Yes), the plasma processing apparatus 1 obtains the parameters derived from the virtual ALD and performs ALD to which the obtained parameters are applied. This ALD is a process in an actual experiment. The ALD performed may be conformal ALD or sub-conformal ALD. ALD is performed according to the process of the following steps S110 to S118.

[0090] The plasma processing apparatus 1 introduces a first gas (precursor) into the plasma processing chamber 10 (step S110 ). Then, the plasma processing apparatus 1 purges the plasma processing chamber 10 to discharge components of the first gas excessively adsorbed on the surface of the object to be processed (step S111 ).

[0091] The plasma processing apparatus 1 introduces the second gas (reaction gas) into the plasma processing chamber 10 to generate plasma of the second gas (step S112 ).

[0092] The control unit 2 acquires the sensing data output from the first sensor 108a during the generation of the plasma, and measures / estimates the plasma state based on the acquired sensing data (step S113). The control unit 2 determines whether the plasma state in the plasma processing chamber 10 is a desired state based on the measurement / estimation result of the plasma state (step S114). When it is determined that the plasma state is not a desired state (S114: "No"), the control unit 2 adjusts control parameters such as the source RF power and the bias RF power (step S115), and returns the process to step S113.

[0093] When it is determined that the plasma state in plasma processing chamber 10 is in a desired state ( S114 : Yes), control unit 2 purges plasma processing chamber 10 to exhaust excess components of the second gas (step S116 ).

[0094] The control unit 2 acquires the sensing data output from the second sensor 108b during the execution of ALD (actual experiment). The control unit 2 measures / estimates the pattern shape of the object on which the protective film is formed by ALD based on the sensing data of the second sensor 108b (step S117). The shape measured / estimated in step S117 may be the shape of each concave portion formed on the surface of the object, or the uniformity of the overall concave shape in the surface of the object.

[0095] The control unit 2 determines whether the ideal shape is obtained based on the measurement / estimation result of the pattern shape (step S118). The control unit 2 determines whether the ideal shape is obtained by measuring / estimating the shape of the protective film formed by ALD based on the sensor data obtained from the second sensor 108b and determining whether the protective film has a desired thickness. If the ideal shape is not obtained (S118: No), the control unit 2 returns the process to step S110 and repeats ALD.

[0096] Furthermore, the control unit 2 may compare the measurement / estimation result of the pattern shape with the set value set for the pattern shape, and may stop subsequent processing or output an alarm based on the comparison result.

[0097] The control unit 2 can also obtain various output data and various measurement data output from the plasma processing apparatus 1 during the execution of ALD at any time, and repeatedly execute the virtual ALD with reference to the obtained data. The control unit 2 can derive parameters to be applied to the actual experiment from the virtual ALD, and apply the parameters to the ALD (actual experiment) repeatedly executed in the plasma processing apparatus 1.

[0098] After executing the virtual ALD in step S108, the control unit 2 executes virtual etching (step S119). The control unit 2 uses the pattern shape obtained by the virtual ALD in step S108 as the initial shape to estimate the shape of the processed object after the etching process through simulation. In addition, the control unit 2 can also determine whether the result of the virtual etching has obtained the desired shape. If it is determined that the desired shape has not been obtained, the control unit 2 returns the process to step S104 or S108 and repeats the virtual experiment (virtual etching and virtual ALD).

[0099] The control unit 2 obtains various parameters used in the virtual etching, and derives the parameters that should be applied to the actual experiment based on the obtained parameters (step S120). The parameters derived in the virtual etching are the same as the parameters derived in step S104. As in step S104, the control unit 2 can derive the parameters that should be applied to the actual experiment by inputting the parameters obtained in step S120 into a learning model of machine learning including deep learning, machine learning, etc., a statistical model, or a model based on a combination thereof. In addition, the control unit 2 can also optimize the parameters that should be applied to the actual experiment so that the consistency between the shape measured / estimated in the actual experiment and the shape predicted in the virtual experiment becomes higher or the process processing time (throughput) is shortened.

[0100] The plasma processing apparatus 1 acquires the parameters derived from the virtual experiment in step S119 and performs etching using the acquired parameters (step S121). This etching process is a process in the actual experiment.

[0101] The control unit 2 acquires the sensing data output from the second sensor 108b during the etching (actual experiment). The control unit 2 measures / estimates the pattern shape of the pattern formed by the etching of step S121 based on the sensing data of the second sensor 108b (step S122). The shape measured / estimated in step S122 may be the shape of each concave portion formed on the surface of the processed body, or the uniformity of the overall concave shape in the surface of the processed body.

[0102] The control unit 2 determines whether the ideal shape is obtained based on the measurement / estimation results of the pattern shape (step S123). The determination method of step S123 is the same as the determination method of step S107.

[0103] If it is determined that the ideal shape has not been obtained (S123: No), the control unit 2 returns the process to step S105. If it is determined that the ideal shape has been obtained (S123: Yes), the control unit 2 ends the process based on this flowchart.

[0104] Furthermore, the control unit 2 may compare the measurement / estimation result of the pattern shape with the set value set for the pattern shape, and may stop subsequent processing or output an alarm based on the comparison result.

[0105] exist Figure 6 and Figure 7 In the flowchart, the structure in which etching and ALD are performed in one plasma processing apparatus 1 is set, but it can also be set as a structure in which etching and ALD are performed using multiple process modules PM1 to PM6. When etching and ALD are performed in one plasma processing apparatus 1, the control unit 2 can measure / estimate the pattern shape and plasma state based on the output of the optical sensor 108 provided in the plasma processing apparatus 1 (in situ). On the other hand, when multiple process modules PM1 to PM6 are used, for example, the observation device OC can be used to measure / estimate the pattern shape (non-in situ), and the first sensor 108a provided in each process module PM1 to PM6 can be used to measure / estimate the plasma state.

[0106] In addition, Figure 6 and Figure 7 In the flowchart, a structure is set to execute virtual etching and virtual ALD in the control unit 2, but it can also be set to a structure in which the required information is handed over between the control unit 2 and the control device MC and the virtual etching and virtual ALD are executed in the control device MC.

[0107] Next, a configuration example of a simulator used in virtual etching and virtual ALD will be described.

[0108] Figure 8 1 is an explanatory diagram for explaining a configuration example of an etching simulator. The etching simulator is composed of, for example, a plasma simulator SIM1, a shape simulator SIM2, and a reaction product simulator SIM3. These simulators SIM1 to SIM3 are simulators based on a particle model.

[0109] The plasma simulator SIM1 obtains the spatial distribution of the reactive species (ions, radicals, etc.) existing in the plasma processing chamber 10 based on the process condition information, and obtains the incident information such as the incident angle and incident energy of the reactive species. Here, the process condition information is the type of reactive gas, gas flow rate, gas mixing ratio, gas pressure, source RF power, bias RF power, etc.

[0110] In the plasma simulator SIM1, for example, the electric field distribution is obtained based on the Poisson equation, the spatial distribution of the reactants is calculated by the particle Monte Carlo method, and the motion of the reactants near the object to be processed is sampled to obtain the incident information such as the incident angle and incident energy of the reactants incident on the object to be processed. Here, in the particle Monte Carlo method, super particles are used to represent the charged particles in the plasma, and the overall behavior of the plasma is simulated by tracking the trajectories of thousands to hundreds of thousands of super particles.

[0111] In the shape simulator SIM2, in addition to the distribution amount and incident information of the reactive species obtained by the plasma simulator SIM1, the information of the pattern shape on the surface of the processed object is also used to obtain the local etching reaction amount and the macro etching reaction amount. In addition, in the plasma simulator SIM1, the distribution amount and incident information of the reactive species are appropriately updated using the local etching reaction amount and the macro etching reaction amount obtained by the shape simulator SIM2.

[0112] In the reaction product simulator SIM3, in addition to using the distribution amount and incident information of the reaction species obtained by the plasma simulator SIM1, the local etching reaction amount and the macroscopic etching reaction amount obtained by the shape simulator SIM2 are also used to calculate the local adhesion amount of the reaction product, and the macroscopic adhesion amount of the reaction product is calculated to calculate the total adhesion amount of the reaction product.

[0113] In the shape simulator SIM2, the space divided by the pattern shape is divided into a grid, and the Monte Carlo method is used to make the reaction species and the generation species fly to the space in a manner according to the incident angle obtained from the plasma simulator SIM1. In addition, it is set so that when the particles collide with the wall surface of the mask, etc., they react with a certain probability. When the amount of the reaction species in the grid becomes above a certain value, the material of the grid part is deleted to correspond to the phenomenon of disappearance due to the progress of etching. In addition, when the amount of the generation species becomes above a certain value, the material corresponding to the generation species (such as a polymer, etc.) is attached to the wall surface to correspond to the deposition reaction. The control unit 2 uses such an etching simulator to repeatedly perform calculations to obtain the etching shape of the processed object.

[0114] exist Figure 8 The structure example of the etching simulator is shown in FIG. 1 , and the structure example of the ALD simulator is the same. That is, the ALD simulator is composed of, for example, a plasma simulator, a shape simulator, and a reaction product simulator. The plasma simulator calculates the distribution amount and incident information of the reaction species, the shape simulator calculates the local and macroscopic deposition reaction amount, and the reaction product simulator calculates the local and macroscopic adhesion amount of the reaction product. The control unit 2 can use the ALD simulator configured in this way to obtain the shape of the protective film formed on the side wall of the recess.

[0115] exist Figure 8 In the embodiment, an etching simulator composed of a plasma simulator SIM1, a shape simulator SIM2, and a reaction product simulator SIM3 is described, but the structure of the simulator is not limited to Figure 8 The control unit 2 can calculate the etching shape and the shape of the protective film using any simulator (model) that can virtually represent the process.

[0116] In the substrate processing system PS according to the present embodiment, actual experiments and virtual experiments are performed in parallel. When the control unit 2 obtains results based on the actual experiments and the virtual experiments, it may update the simulator (model) used for the virtual experiments in a manner suitable for the results based on the actual experiments.

[0117] Fig. 9 is a flowchart illustrating the updating process of the model. Figure 6 and Figure 7 The process shown performs actual experiments and virtual experiments to obtain the results of the actual experiments and the results of the virtual experiments (steps S201 and S202).

[0118] The control unit 2 calculates the difference between the result of the actual experiment and the result of the virtual experiment (step S203), and determines whether the simulator (model) needs to be updated (step S204). If the calculated difference is greater than the set value, the control unit 2 determines to update the simulator (model) (S204: "Yes"), and updates the simulator (model) (step S205). Specifically, the control unit 2 changes at least one of the parameters constituting the simulator from the value before the update to the value after the update.

[0119] The control unit 2 uses the updated simulator (model) to perform the virtual experiment including virtual etching and virtual ALD again (step S206), and returns the process to step S202. The control unit 2 updates the model as appropriate by repeating the processes from step S202 to step S206.

[0120] When the difference calculated in step S203 is smaller than the threshold value, the control unit 2 determines that updating is not necessary (S204: No), and ends the processing based on this flowchart.

[0121] Fig.10 It is an explanatory diagram for explaining the film thickness of the protective film formed by the substrate processing method according to the embodiment. Fig.10(A) is a schematic diagram of the object to be processed used in the experiment. The object to be processed includes an etching target film EL1 and a mask MA. A recess having an opening OP is formed in the stack of the etching target film EL1 and the mask MA, and a protective film PF is formed on the side wall of the recess. In this embodiment, the opening size CD at an arbitrary position of the space surrounded by the side wall of the recess (protective film PF) is measured.

[0122] Fig.10 (B) is a graph showing the measurement results. The vertical axis of the graph represents the depth of the recess, and the horizontal axis of the graph represents the opening size at any position of the space surrounded by the side wall of the recess. Reference Example 1 shows the experimental results obtained by performing only etching without ALD, and Reference Example 2 shows the experimental results obtained by performing both etching and ALD. In contrast, the embodiment shows the experimental results when the process conditions are derived by virtual etching and virtual ALD according to the substrate processing method involved in this embodiment and applied to the actual experiment.

[0123] In Reference Example 1, the opening size CD decreases as the depth increases from the position of 0.4 μm in depth. The maximum value of the opening size CD in the range shown in the graph is 54.1 nm, and the minimum value is 46.1 nm, so the difference is 8.0 nm.

[0124] In Reference Example 2, the opening size CD increases at a depth of 0.4 μm to 1.2 μm. The maximum value of the opening size CD in the range shown in the graph is 49.2 nm, and the minimum value is 42.2 nm, so the difference is 7.0 nm.

[0125] In contrast, in the embodiment, it is known that a protective film is formed with a substantially constant film thickness regardless of the depth. The maximum value of the opening size CD within the range shown in the graph is 45.6 nm, and the minimum value is 40.0 nm, so the difference is 5.6 nm. That is, it can be qualitatively shown that a good opening shape is obtained compared to Reference Examples 1 and 2.

[0126] As described above, in this embodiment, the process conditions derived from the virtual experiment (virtual etching and virtual ALD) are applied to perform actual experiments, so that a good opening shape can be guided. In addition, by updating the model based on the experimental results obtained in the actual experiment and the experimental results obtained in the virtual experiment, the number of trials of the actual experiment can be reduced, and a good opening shape can be guided more efficiently.

[0127] (Implementation Method 2)

[0128] In Embodiment 2, a configuration for deriving parameters to be applied to actual experiments using a reinforcement learning method is described.

[0129] The overall structure of the substrate processing system and the device structure of each device are the same as those in Embodiment 1, and thus the description thereof is omitted.

[0130] In this embodiment, a reinforcement learning algorithm is used to derive parameters that should be applied to actual experiments. Fig.11 This is a schematic diagram of the reinforcement learning algorithm. The reinforcement learning algorithm is an algorithm that deals with the problem of making an intelligent agent placed in a certain environment observe the current state of the observed object and decide what action to take. Next, DQN (Deep Q-Network) as a reinforcement learning method is explained.

[0131] The learning model in reinforcement learning is learned as the current state s of the input observation object t In the case of , the value of the action value function (Q value) is output for each of the actions a1, a2, ..., an (n is an integer greater than 2) that can be taken. DQN is a method of performing reinforcement learning by approximating the action value function through a neural network.

[0132] In this embodiment, a learning model MD is expressed using a neural network that approximates an action-value function, and reinforcement learning is performed in a manner that outputs information on the value when selecting parameters to be applied to actual experiments according to the current state of the processed object.

[0133] The state s is input to the learning model MD t For example, the data of the shape measured / estimated in the actual experiment. The learning model MD is for the current state s t The action value function Q(s) is output for each of the possible actions a1, a2, ..., an (n is an integer greater than 2). t ,a1),Q(s t ,a2),…,Q(s t , an). The value of the action value function represents the value of the action value function in state s t The expected value of the future benefit that can be obtained if action a is selected is also called Q value. That is, the value of the action value function (Q value) does not represent short-term rewards, but represents the value in the long term. In this embodiment, action a is equivalent to performing an actual experiment according to the selected parameters.

[0134] The agent refers to the Q value output from the learning model MD for each action to choose the state s t The action a with the highest Q value among the actions a1, a2, ..., an that can be taken t The environment is based on the action a selected t And is updated, the next state s t+1In this embodiment, the agent is the control unit 2, and the environment is a simulator for conducting virtual experiments.

[0135] The agent takes action a t The next state s generated by the choice t+1 , to obtain reward r from the environment t+1 . Reward t+1 For example, the ratio of the shape of the concave portion measured / estimated by monitoring in the actual experiment to the shape of the concave portion predicted by the virtual experiment. t+1 It can also be the processing time.

[0136] The agent repeatedly tries and fails to learn actions that maximize future rewards (benefits). Specifically, the agent uses the state s t , status t+1 And for the last action a t Rewards t+1 , to sequentially update the learning model MD based on the following expression (Japanese: formula) (1).

[0137] Q(s t ,a t )←Q(s t ,a t )+α(r t+1 +γ·maxQ(s t+1 ,a t+1 )-Q(s t ,a t )}

[0138] ……(1)

[0139] Here, α is the learning coefficient, γ is the discount rate, and r t+1 is as an action t The learning coefficient α is a parameter that determines the speed of learning and satisfies the relationship of 0<α<1. The discount rate γ is a parameter that indicates the degree of discount for the evaluation of the future state and satisfies the relationship of 0<γ<1.

[0140] In Q-learning, the model parameters of the learning model MD are learned using the error back propagation method or the like so that the second term on the right side of the expression (1) becomes zero. This means that in the case of action a t The state s t Transformed to state s t+1 In the case of t The Q value is close to the next state s t+1 The value of the case where the Q value is the highest.

[0141] The agent repeatedly updates the learning model MD until the specified termination condition is met. By repeatedly updating, the learning model MD is learned to make the reward r t+1 The end condition is appropriately set to, for example, when a predetermined number of updates have been performed, when the shape of the concave portion of the processed object has approached the target shape, when the processed object can no longer be reduced, and the like.

[0142] When the learning model MD is obtained by the above-mentioned learning algorithm, the control unit 2 can use the learning model MD to derive parameters to be applied to the actual experiment. Specifically, the control unit 2 inputs the current state s of the observed object into the learned learning model MD. t (data of shapes measured / estimated in actual experiments) and performs calculations by the learning model MD. As a result of the calculations by the learning model MD, Q values ​​can be obtained for each of the actions a1, a2, ..., an that can be taken. The control unit 2 can derive parameters that should be applied to the actual experiment by selecting the action with the highest Q value.

[0143] As described above, in Embodiment 2, reinforcement learning can be used to derive parameters that should be applied to actual experiments performed in the plasma processing apparatus 1. In addition, in this embodiment, the structure for deriving parameters that should be applied to actual experiments in one plasma processing apparatus 1 is described, but it goes without saying that the parameters derived for one plasma processing apparatus 1 can also be applied to one or more other plasma processing apparatuses.

[0144] In this embodiment, a learning algorithm based on Q-learning is described as an example, but the method of generating the learning model MD is not limited to Q-learning. For example, any reinforcement learning algorithm such as TD learning (Temporal Difference Learning), policy gradient method (Policy gradients), SARSA (State-Action-Reward-State-Action), Actor-critic, etc. can be used.

[0145] (Implementation method 3)

[0146] In the first embodiment, the virtual experiment and parameter export are performed in the control unit 2 that controls the operation of the plasma processing apparatus 1 , but the virtual experiment and parameter export may be performed in an external server device that is communicably connected to the control unit 2 .

[0147] In the third embodiment, a configuration in which a virtual experiment and parameter derivation are performed in an external server device will be described.

[0148] Fig.12 1 is a schematic diagram showing a first configuration example of a plasma processing system in Embodiment 3. The plasma processing system in Embodiment 3 includes a plasma processing apparatus 1, a control unit 2, and a server device 3. The configurations of the plasma processing apparatus 1 and the control unit 2 are the same as those in Embodiment 1, and thus description thereof is omitted.

[0149] The server device 3 is a computer connected to the control unit 2 in a communicative manner via the communication network NW, and includes a processing unit 3a, a storage unit 3b, a communication unit 3c, etc. The processing unit 3a includes a CPU, a ROM, a RAM, etc., and performs a virtual experiment that simulates the process (actual experiment) performed in the plasma processing device 1, and derives parameters that should be applied to the actual experiment. The storage unit 3b includes storage devices such as HDD and SDD. The storage unit 3b includes a simulator for virtual experiments used to simulate the process (actual experiment) performed in the substrate processing system PS. The simulators for virtual experiments include a plasma simulator, a reaction product simulator, a shape simulator, etc. The communication unit 3c includes a communication interface for communicating with the control unit 2 via the communication network NW.

[0150] The server device 3 acquires shape data measured / estimated by the control unit 2 when a process (actual experiment) is performed in the plasma processing device 1 via the communication network NW. The server device 3 uses the shape data acquired from the plasma processing device 1 as initial values ​​to perform virtual etching and virtual ALD, and estimates the shape of the processed object after the processing through simulation. The server device 3 acquires various parameters used in the virtual etching and virtual ALD, and derives parameters to be applied to the actual experiment based on the acquired parameters.

[0151] The server device 3 may also optimize the parameters to be applied to the actual experiment so that the consistency between the shape measured / estimated in the actual experiment and the shape predicted in the virtual experiment becomes higher or the process time (throughput) is shortened. The server device 3 may also use a reinforcement learning method to derive the parameters to be applied to the actual experiment.

[0152] The above-mentioned processing executed by the server device 3 is the same as the virtual experiment process described in Embodiments 1 and 2, and therefore the detailed description thereof will be omitted.

[0153] The server device 3 transmits the derived parameters to the control unit 2 via the communication network NW. The plasma processing apparatus 1 executes etching and ALD using the parameters received by the control unit 2 from the server device 3. These etching and ALD are processes in actual experiments.

[0154] The above-mentioned processing performed by the plasma processing apparatus 1 is the same as the actual experimental process described in Embodiments 1 and 2, and therefore the detailed description thereof is omitted.

[0155] As described above, in Embodiment 3, a virtual experiment (virtual etching, virtual ALD) can be performed in the server device 3 connected to the control unit 2 in a communicative manner, and parameters to be applied to the plasma processing device 1 can be derived and fed back to the plasma processing device 1.

[0156] Fig.13 1 is a schematic diagram showing a second configuration example of a plasma processing system in Embodiment 3. In addition to the plasma processing apparatus 1, the control unit 2, and the server device 3 described above, the plasma processing system in Embodiment 3 further includes plasma processing apparatuses 1-1, 1-2, ..., 1-n, and control units 2-1, 2-2, ..., 2-n. The control units 2-1, 2-2, ..., 2-n respectively control the operations of the plasma processing apparatuses 1-1, 1-2, ..., 1-n, and are respectively connected to a communication network NW.

[0157] and Fig.12 Similarly to the example, the server device 3 acquires data obtained from the plasma processing device 1 to perform a virtual experiment (virtual etching, virtual ALD), and derives parameters that should be applied to the plasma processing device 1. The server device 3 sends the derived parameters to the control unit 2 that controls the plasma processing device 1, and to the control units 2-1, 2-2, ..., 2-n that control the plasma processing devices 1-1, 1-2, ..., 1-n. The plasma processing device 1 performs etching and ALD using the parameters received by the control unit 2 from the server device 3. Similarly, the plasma processing devices 1-1, 1-2, ..., 1-n perform etching and ALD using the parameters received by the control units 2-1, 2-2, ..., 2-n from the server device 3.

[0158] As described above, in the second configuration example, parameters to be applied to the plasma processing apparatus 1 can be derived and fed back to the plasma processing apparatus 1, and can also be fed back to other plasma processing apparatuses 1-1, 1-2, . . . , 1-n.

[0159] The embodiments disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is indicated by the claims rather than the above meaning, and is intended to include all modifications within the meaning and scope equivalent to the claims.

[0160] In the embodiment, an application example for a capacitively coupled plasma processing device 1 is described, but the application is not limited to the capacitively coupled type, and can also be applied to any type of plasma processing device including an inductively coupled plasma (ICP), a radial line slot antenna (RLSA), an electron cyclotron resonance plasma (ECR), and a helicon wave plasma (HWP). In addition, CVD (Chemical Vapor Deposition) can also be used instead of ALD.

[0161] In addition, the present disclosure includes the following embodiments.

[0162] (Note 1)

[0163] A control method for a substrate processing device comprises the following steps:

[0164] a) locally etching the object to be processed to form a recessed portion in the object to be processed;

[0165] b) forming a protective film on the side wall of the recessed portion formed in the object to be processed;

[0166] c) further etching the object to be processed having the recess and the protective film formed therein;

[0167] d) repeating b) and c);

[0168] e) monitoring the processed object obtained through at least one of a) to d);

[0169] f) conducting a virtual experiment simulating the steps a) to d);

[0170] g) deriving a parameter to be applied to at least one of a) to d) based on the monitoring result of the treated object and the result of the virtual experiment; and

[0171] h) Implementing at least one of a) to d) applying the derived parameters.

[0172] (Note 2)

[0173] According to the control method of the substrate processing device described in Supplement 1,

[0174] The method further includes the step of i) updating the model for deriving the parameter based on a difference between the shape of the recessed portion estimated by monitoring the object to be processed and the shape of the recessed portion predicted by the virtual experiment.

[0175] (Note 3)

[0176] A control method for a substrate processing apparatus according to Appendix 1 or 2, wherein:

[0177] The step of forming the protective film includes at least one selected from the group consisting of CVD (Chemical Vapor Deposition), ALD (Atomic Layer Deposition), and unsaturated ALD.

[0178] (Note 4)

[0179] A control method for a substrate processing apparatus according to Appendix 1 or 2, wherein:

[0180] The process of forming the protective film includes the following steps:

[0181] supplying a gas containing a precursor to the object to be processed so that the precursor is adsorbed on at least a side wall of the recess; and

[0182] A reaction gas that reacts with the precursor is supplied to the object to be processed, and the protective film is formed by a reaction between the reaction gas and the precursor adsorbed in the concave portion.

[0183] (Note 5)

[0184] A control method for a substrate processing apparatus according to any one of Supplementary Notes 1 to 4, wherein:

[0185] The parameters include film forming conditions of the protective film.

[0186] (Note 6)

[0187] According to the control method of the substrate processing apparatus described in Supplementary Note 5,

[0188] The film forming conditions include at least one of a film forming method, a number of cycles, time, gas species, dilution, temperature, and RF power.

[0189] (Note 7)

[0190] A control method for a substrate processing apparatus according to any one of Supplementary Notes 1 to 6, wherein:

[0191] The parameters include etching conditions for the object to be processed.

[0192] (Note 8)

[0193] According to the control method of the substrate processing device described in Supplementary Note 7, wherein:

[0194] The etching conditions include at least one of RF power, processing time, gas species, a mixing ratio of gases, and temperature.

[0195] (Note 9)

[0196] A control method for a substrate processing apparatus according to any one of Supplementary Notes 1 to 8, wherein:

[0197] The a) to d) or the a) to e) are performed in the same chamber.

[0198] (Note 10)

[0199] A control method for a substrate processing apparatus according to any one of Supplementary Notes 1 to 8, wherein:

[0200] The above a) to e) are performed in at least two or more chambers.

[0201] (Note 11)

[0202] According to the control method of the substrate processing device described in Supplement 1,

[0203] The method further includes the step of j) stopping the treatment or outputting an alarm based on a comparison result of the shape of the recessed portion of the treated object estimated by monitoring and a set value set in relation to the shape.

[0204] (Note 12)

[0205] The control method of the substrate processing device according to Supplement 1 further includes the following steps:

[0206] k) generating a learning model for deriving the parameters by reinforcement learning, wherein the reinforcement learning takes the consistency rate between the shape of the concave portion of the processed object estimated by monitoring and the shape of the concave portion predicted by the virtual experiment or the process time as a reward; and

[0207] 1) applying the parameters derived using the learning model to substrate processing in one or more substrate processing apparatuses including the substrate processing apparatus.

[0208] (Note 13)

[0209] A control method for a substrate processing device, wherein:

[0210] The substrate processing apparatus is configured to perform a process including the following steps:

[0211] a) locally etching the object to be processed to form a recessed portion in the object to be processed;

[0212] b) forming a protective film on the side wall of the recessed portion formed in the object to be processed;

[0213] c) further etching the object to be processed having the recess and the protective film formed therein; and

[0214] d) repeating b) and c),

[0215] The control method comprises the following steps:

[0216] i) acquiring first data representing the shape of the processed object after any one of the processes a) to d);

[0217] ii) performing a virtual experiment simulating steps a) to d) on the object to be processed, and acquiring second data representing the shape of the object to be processed after the virtual experiment; and

[0218] iii) deriving a parameter that should be applied to at least one of a) to d) based on the first data and the second data.

[0219] (Note 14)

[0220] A substrate processing system includes a substrate processing device and a control device, wherein:

[0221] The substrate processing device performs the following steps:

[0222] a) locally etching the object to be processed to form a recessed portion in the object to be processed;

[0223] b) forming a protective film on the side wall of the recessed portion formed in the object to be processed;

[0224] c) further etching the object to be processed having the recess and the protective film formed therein; and

[0225] d) repeating b) and c),

[0226] The control device comprises:

[0227] A monitoring unit that monitors the object to be processed obtained by at least one of a) to d);

[0228] A simulation unit, which performs virtual experiments simulating the above a) to d); and

[0229] a deriving unit that derives a parameter to be applied to at least one of a) to d) based on the monitoring result of the treated object and the result of the virtual experiment,

[0230] The control device causes the substrate processing apparatus to execute at least one of a) to d) to which the derived parameter is applied.

[0231] (Note 15)

[0232] A substrate processing system according to Appendix 14, wherein:

[0233] The control device stops the treatment or outputs an alarm based on a comparison result of the shape of the recessed portion estimated by monitoring the object to be treated and a set value set for the shape.

[0234] (Note 16)

[0235] A substrate processing system according to Appendix 14, wherein:

[0236] The control device generates a learning model for deriving the parameters by reinforcement learning, wherein the reinforcement learning takes the consistency rate between the shape of the concave portion of the processed object estimated by monitoring and the shape of the concave portion predicted by the virtual experiment or the process processing time as a reward,

[0237] The control apparatus applies the parameters derived using the learning model to substrate processing in one or more substrate processing apparatuses including the substrate processing apparatus.

[0238] (Note 17)

[0239] A substrate processing system includes a substrate processing device and a control device, wherein:

[0240] The substrate processing apparatus is configured to perform a process including the following steps:

[0241] a) locally etching the object to be processed to form a recessed portion in the object to be processed;

[0242] b) forming a protective film on the side wall of the recessed portion formed in the object to be processed;

[0243] c) further etching the object to be processed having the recess and the protective film formed therein; and

[0244] d) repeating b) and c),

[0245] The control device is configured to execute a process including the following steps:

[0246] i) acquiring first data representing the shape of the processed object after any one of the processes a) to d);

[0247] ii) performing a virtual experiment simulating steps a) to d) on the object to be processed, and acquiring second data representing the shape of the object to be processed after the virtual experiment; and

[0248] iii) deriving a parameter that should be applied to at least one of a) to d) based on the first data and the second data.

[0249] Description of Reference Numerals

[0250] 1: plasma processing device; 2: control unit; 2a: computer; 2a1: processing unit; 2a2: storage unit; 2a3: communication interface; 10: plasma processing chamber.

Claims

1. A control method for a substrate processing device, comprising the following steps: a) locally etching the object to be processed to form a recessed portion in the object to be processed; b) forming a protective film on the side wall of the recessed portion formed in the object to be processed; c) further etching the object to be processed having the recess and the protective film formed therein; d) repeating b) and c); e) monitoring the processed object obtained through at least one of a) to d); f) conducting a virtual experiment simulating the steps a) to d); g) deriving a parameter to be applied to at least one of a) to d) based on the monitoring result of the treated object and the result of the virtual experiment; as well as h) Implementing at least one of a) to d) applying the derived parameters.

2. The control method of a substrate processing apparatus according to claim 1, wherein: The method further includes the step of i) updating the model for deriving the parameter based on a difference between the shape of the recessed portion estimated by monitoring the object to be processed and the shape of the recessed portion predicted by the virtual experiment.

3. The control method of a substrate processing apparatus according to claim 1, wherein: The process of forming the protective film includes at least one selected from the group consisting of chemical vapor deposition (CVD), atomic layer deposition (ALD), and unsaturated ALD.

4. The control method of a substrate processing apparatus according to claim 1, wherein: The process of forming the protective film includes the following steps: supplying a gas containing a precursor to the object to be processed so that the precursor is adsorbed on at least a side wall of the recess; and A reaction gas that reacts with the precursor is supplied to the object to be processed, and the protective film is formed by a reaction between the reaction gas and the precursor adsorbed in the concave portion.

5. The control method of a substrate processing apparatus according to claim 1, wherein: The parameters include film forming conditions of the protective film.

6. The control method of a substrate processing apparatus according to claim 5, wherein: The film forming conditions include at least one of a film forming method, a number of cycles, a time, a gas species, a dilution, a temperature, and a radio frequency power.

7. The control method of a substrate processing apparatus according to claim 1, wherein: The parameters include etching conditions for the object to be processed.

8. The control method of a substrate processing apparatus according to claim 7, wherein: The etching conditions include at least one of radio frequency power, processing time, gas species, a mixing ratio of gases, and temperature.

9. The control method of a substrate processing apparatus according to any one of claims 1 to 8, wherein: The a) to d) or the a) to e) are performed in the same chamber.

10. The control method of a substrate processing apparatus according to any one of claims 1 to 8, wherein: The above a) to e) are performed in at least two or more chambers.

11. The control method of a substrate processing apparatus according to claim 1, wherein: The method further includes the step of j) stopping the treatment or outputting an alarm based on a result of comparing the shape of the recessed portion of the object to be treated estimated by monitoring with a set value set in relation to the shape.

12. The control method of the substrate processing apparatus according to claim 1, further comprising the following steps: k) generating a learning model for deriving the parameters by reinforcement learning, wherein the reinforcement learning takes the consistency rate between the shape of the concave portion of the processed object estimated by monitoring and the shape of the concave portion predicted by the virtual experiment or the process time as a reward; and 1) applying the parameters derived using the learning model to substrate processing in one or more substrate processing apparatuses including the substrate processing apparatus.

13. A method for controlling a substrate processing device, wherein: The substrate processing apparatus is configured to perform a process including the following steps: a) locally etching the object to be processed to form a recessed portion in the object to be processed; b) forming a protective film on the side wall of the recessed portion formed in the object to be processed; c) further etching the object to be processed having the recess and the protective film formed therein; and d) repeating b) and c), The control method comprises the following steps: i) acquiring first data representing the shape of the processed object after any one of the processes a) to d); ii) performing a virtual experiment simulating steps a) to d) on the object to be processed, and acquiring second data representing the shape of the object to be processed after the virtual experiment; and iii) deriving a parameter that should be applied to at least one of a) to d) based on the first data and the second data.

14. A substrate processing system, comprising a substrate processing device and a control device, wherein: The substrate processing device performs the following steps: a) locally etching the object to be processed to form a recessed portion in the object to be processed; b) forming a protective film on the side wall of the recessed portion formed in the object to be processed; c) further etching the object to be processed having the recess and the protective film formed therein; and d) repeating b) and c), The control device comprises: A monitoring unit that monitors the object to be processed obtained by at least one of a) to d); A simulation unit, which performs virtual experiments simulating the above a) to d); as well as a deriving unit that derives a parameter to be applied to at least one of a) to d) based on the monitoring result of the treated object and the result of the virtual experiment, The control device causes the substrate processing apparatus to execute at least one of a) to d) to which the derived parameter is applied.

15. The substrate processing system according to claim 14, wherein: The control device stops the treatment or outputs an alarm based on a comparison result of the shape of the recessed portion estimated by monitoring the object to be treated and a set value set for the shape.

16. The substrate processing system according to claim 14, wherein: The control device generates a learning model for deriving the parameters by reinforcement learning, wherein the reinforcement learning takes the consistency rate between the shape of the concave portion of the processed object estimated by monitoring and the shape of the concave portion predicted by the virtual experiment or the process processing time as a reward, The control apparatus applies the parameters derived using the learning model to substrate processing in one or more substrate processing apparatuses including the substrate processing apparatus.

17. A substrate processing system, comprising a substrate processing device and a control device, wherein: The substrate processing apparatus is configured to perform a process including the following steps: a) locally etching the object to be processed to form a recessed portion in the object to be processed; b) forming a protective film on the side wall of the recessed portion formed in the object to be processed; c) further etching the object to be processed having the recess and the protective film formed therein; as well as d) repeating b) and c), The control device is configured to execute a process including the following steps: i) acquiring first data representing the shape of the processed object after any one of the processes a) to d); ii) performing a virtual experiment simulating steps a) to d) on the object to be processed, and acquiring second data representing the shape of the object to be processed after the virtual experiment; as well as iii) deriving a parameter that should be applied to at least one of a) to d) based on the first data and the second data.

Images