Information processing method, information processing device, and computer program

Abstract

Provided are an information processing method, an information processing device, and a computer program.　The present invention acquires, for each of a plurality of conditions obtained by changing one or more setting values in a plasma treatment recipe, time series data measured by a specific sensor when plasma treatment is executed, calculates, for each of the plurality of conditions, the average value of the time series data in a specific step section of the plasma treatment recipe, derives a correlation between the setting value and the calculated average value, compares, for each of the plurality of conditions, the calculated average value and a predicted value predicted from the correlation, and determines the stability of the plasma treatment recipe on the basis of the comparison result between the average value and the predicted value.

Description

Information Processing Method, Information Processing Apparatus, and Computer Program

[0001] The present disclosure relates to an information processing method, an information processing apparatus, and a computer program.

[0002] Conventionally, in the field of various manufacturing processes, anomaly detection techniques for detecting anomalies occurring in various processes from time-series data obtained during the manufacturing of an object are known (see Patent Document 1).

[0003] For example, in a semiconductor manufacturing process, by using a pre-generated anomaly detection model to monitor a time-series data group measured during the processing of a wafer, the presence or absence of an anomaly, the degree of anomaly, etc. are determined.

[0004] Japanese Unexamined Patent Application Publication No. 2021 - 86571

[0005] The present disclosure provides an information processing method, an information processing apparatus, and a computer program for determining the stability of a plasma processing recipe.

[0006] The information processing method of the present disclosure acquires time-series data measured by a specific sensor when plasma processing is performed for each of a plurality of conditions obtained by changing one or more set values in a plasma processing recipe, calculates the average value of the time-series data in a specific step interval of the plasma processing recipe for each of the plurality of conditions, derives the correlation between the set value and the calculated average value, compares the calculated average value with a predicted value predicted from the correlation for each of the plurality of conditions, and causes a computer to execute a process of determining the stability of the plasma processing recipe based on the comparison result between the average value and the predicted value.

[0007] According to the present disclosure, the stability of a plasma processing recipe can be determined.

[0008] This is a diagram illustrating an example configuration of a plasma processing system. This is a diagram illustrating an example configuration of a capacitively coupled plasma processing apparatus. This is an explanatory diagram illustrating an overview of the processing performed by the processing unit. This is an explanatory diagram illustrating the details of score checking. This is an explanatory diagram illustrating how to set the values. This is a graph showing an example of log data. This is a graph showing a comparison example between measured values ​​and predicted values. This is a chart showing a comparison example with reference values. This is a graph showing another example of log data. This is a graph showing an example of calculating the difference between measured values ​​and moving averages. This is a flowchart illustrating the procedure of the processing performed by the processing unit. This is a schematic diagram showing the first example. This is a schematic diagram showing the second example.

[0009] An embodiment will be described below with reference to the drawings. In this description, the same elements or elements having the same function will be denoted by the same reference numeral, and redundant descriptions will be omitted.

[0010] Figure 1 is a diagram illustrating an example configuration of a plasma processing system. In one embodiment, the plasma processing system includes a plasma processing apparatus 1 and a control unit 2. The plasma processing system is an example of a substrate processing system, and the plasma processing apparatus 1 is an example of a substrate processing apparatus. The plasma processing apparatus 1 includes a plasma processing chamber 10, a substrate support unit 11, and a plasma generation unit 12. The plasma processing chamber 10 has a plasma processing space. The plasma processing chamber 10 also has at least one gas supply port for supplying at least one processing gas to the plasma processing space, and at least one gas outlet for discharging gas from the plasma processing space. The gas supply port is connected to a gas supply unit 20, which will be described later, and the gas outlet is connected to an exhaust system 40, which will be described later. The substrate support unit 11 is located in the plasma processing space and has a substrate support surface for supporting a substrate.

[0011] The plasma generation unit 12 is configured to generate plasma from at least one processing gas supplied into the plasma processing space. The plasma formed in the plasma processing space may be capacitively coupled plasma (CCP), inductively coupled plasma (ICP), ECR (Electron Cyclotron Resonance) plasma, helicon wave excited plasma (HWP), or surface wave plasma (SWP), etc. Various types of plasma generation units, including AC (Alternating Current) plasma generation units and DC (Direct Current) plasma generation units, may also be used. In one embodiment, the AC signal (AC power) used in the AC plasma generation unit has a frequency in the range of 100 kHz to 10 GHz. Therefore, the AC signal includes an RF (Radio Frequency) signal and a microwave signal. In one embodiment, the RF signal has a frequency in the range of 100 kHz to 150 MHz.

[0012] The control unit 2 processes computer-executable instructions that cause the plasma processing apparatus 1 to perform the various processes described herein. The control unit 2 may be configured to control the elements of the plasma processing apparatus 1 to perform the various processes described herein. In one embodiment, part or all of the control unit 2 may be included in the plasma processing apparatus 1. The control unit 2 is implemented, for example, by a computer 2a. The control unit 2 may include a processing unit 2a1, a storage unit 2a2, and a communication interface 2a3. The functions realized by the processing unit 2a1 described in this disclosure may be implemented in a circuit or processing circuit, including a general-purpose processor, an application-specific processor, integrated circuits, ASICs (Application Specific Integrated Circuits), a CPU (Central Processing Unit), a conventional circuit, and / or a combination thereof, programmed to realize the described functions. The processor is considered to be a circuit or processing circuit, including transistors and other circuits. The processor may be a programmed processor that executes a program stored in the storage unit 2a2. This program may be pre-stored in the storage unit 2a2 and may be retrieved via a medium M when needed. The acquired program is stored in the storage unit 2a2 and read from the storage unit 2a2 and executed by the processing unit 2a1. The medium may be various storage media readable by the computer 2a, or it may be a communication line connected to the communication interface 2a3. The storage unit 2a2 may include RAM (Random Access Memory), ROM (Read Only Memory), HDD (Hard Disk Drive), SSD (Solid State Drive), or a combination thereof. The communication interface 2a3 may communicate with the plasma processing device 1 via a communication line such as a LAN (Local Area Network).In this disclosure, circuits, units, and means are hardware programmed to perform or configured to perform the functions described. Such hardware may be any hardware described in this disclosure, or any hardware known to be programmed to perform or execute the functions described. If such hardware is a processor that is considered to be a type of circuit, such circuit, means, or unit is a combination of hardware and software used to constitute such hardware and / or processor.

[0013] The following describes an example configuration of a capacitively coupled plasma processing apparatus as an example of a plasma processing apparatus 1. Figure 2 is a diagram illustrating an example configuration of a capacitively coupled plasma processing apparatus.

[0014] The capacitively coupled plasma processing apparatus 1 includes a plasma processing chamber 10, a gas supply unit 20, a power supply system 30, and an exhaust system 40. The plasma processing apparatus 1 also 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 located inside the plasma processing chamber 10. The shower head 13 is located 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 walls 10a of the plasma processing chamber 10, and the substrate support unit 11. The plasma processing chamber 10 is grounded. The shower head 13 and the substrate support unit 11 are electrically insulated from the housing of the plasma processing chamber 10.

[0015] The substrate support portion 11 includes a main body portion 111 and a ring assembly 112. The main body portion 111 has a central region 111a for supporting the substrate W and an annular region 111b for supporting the ring assembly 112. A wafer is an example of a substrate W. The annular region 111b of the main body portion 111 surrounds the central region 111a of the main body portion 111 in a plan view. The substrate W is placed on the central region 111a of the main body portion 111, and the ring assembly 112 is placed on the annular region 111b of the main body portion 111 so as to surround the substrate W on the central region 111a of the main body portion 111. Therefore, the central region 111a is also called the substrate support surface for supporting the substrate W, and the annular region 111b is also called the ring support surface for supporting the ring assembly 112.

[0016] In one embodiment, the main body 111 includes a base 1110 and an electrostatic chuck 1111. The base 1110 includes a conductive member. The conductive member of the base 1110 can function as a lower electrode. The electrostatic chuck 1111 is placed on the base 1110. The electrostatic chuck 1111 includes a ceramic member 1111a and an electrostatic chuck electrode 1111b placed within the ceramic member 1111a. The electrostatic chuck electrode 1111b is also called a clamping electrode. In one embodiment, the electrostatic chuck electrode 1111b is electrically connected or coupled to a chuck power supply. The chuck power supply may be a DC power supply or an AC power supply. The ceramic member 1111a has a central region 111a. In one embodiment, the ceramic member 1111a also has an annular region 111b. Furthermore, other members surrounding the electrostatic chuck 1111, such as an annular electrostatic chuck or an annular insulating member, may have an annular region 111b. In this case, the ring assembly 112 may be placed on the annular electrostatic chuck or the annular insulating member, or it may be placed on both the electrostatic chuck 1111 and the annular insulating member. In addition, at least one bias electrode, which is electrically connected or coupled to the power supply 31 and / or power supply 32 described later, may be placed inside the ceramic member 1111a. In this case, at least one bias electrode functions as a lower electrode. Also, the conductive member of the base 1110 and the bias electrode inside the ceramic member 1111a may function as multiple lower electrodes. In one embodiment, the first voltage generation unit 32a, which functions as a voltage pulse generation unit described later, is electrically connected or coupled to the bias electrode inside the ceramic member 1111a, and the first RF generation unit 31a, described later, is electrically connected or coupled to the conductive member of the base 1110. Furthermore, the electrostatic chuck electrode 1111b may function as a lower electrode. Therefore, the substrate support portion 11 includes at least one lower electrode.

[0017] The ring assembly 112 includes one or more annular members. In one embodiment, the one or more annular members include one or more edge rings and at least one covering ring. The edge rings are formed of a conductive or insulating material, and the covering rings are formed of an insulating material.

[0018] The substrate support section 11 may also include a temperature control module configured to adjust at least one of the electrostatic chuck 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 through the flow path 1110a. In one embodiment, the flow path 1110a is formed within the base 1110, and one or more heaters are arranged within the ceramic member 1111a of the electrostatic chuck 1111. The substrate support section 11 may also include a heat transfer gas supply section configured to supply heat transfer gas to the gap between the back surface of the substrate W and the central region 111a.

[0019] The showerhead 13 is configured to introduce at least one processing gas from the gas supply unit 20 into the plasma processing space 10s. The showerhead 13 has at least one gas supply port 13a, at least one gas diffusion chamber 13b, and a plurality of gas inlet ports 13c. The processing gas supplied to the gas supply port 13a passes through the gas diffusion chamber 13b and is introduced into the plasma processing space 10s through the plurality of gas inlet ports 13c. The showerhead 13 also includes at least one upper electrode. In addition to the showerhead 13, the gas introduction unit may also include one or more side gas injectors (SGIs) attached to one or more openings formed in the side wall 10a.

[0020] 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 to the shower head 13 from a corresponding gas source 21 via a corresponding flow controller 22. Each flow controller 22 may include, for example, a mass flow controller or a pressure-controlled flow controller. Furthermore, the gas supply unit 20 may include at least one flow modulation device that modulates or pulses the flow rate of at least one processing gas.

[0021] The power supply system 30 includes a power supply 31 that is electrically connected to or coupled to the plasma processing chamber 10. In one embodiment, the power supply 31 is electrically connected to or coupled to the plasma processing chamber 10 via at least one impedance matcher. The impedance matcher may be a mechanically controlled matcher or an electronically controlled matcher. The 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. This generates plasma from at least one processing gas supplied to the plasma processing space 10s. Therefore, the power supply 31 can function as at least part of the plasma generation unit 12. In addition, by supplying a bias RF signal to at least one lower electrode, a bias potential is generated on the substrate W, and ionic components in the formed plasma can be drawn into the substrate W.

[0022] The power supply 31 includes a first RF generation unit 31a and a second RF generation unit 31b. The first RF generation unit 31a is electrically connected or coupled to at least one lower electrode and / or at least one upper electrode and is configured to generate a source RF signal (source RF power) to generate plasma in the plasma processing space 10s. In one embodiment, the first RF generation unit 31a is electrically connected or coupled to at least one lower electrode and / or at least one upper electrode via at least one impedance matcher. 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 generation unit 31a may be configured to generate a plurality of source RF signals having different frequencies. One or more generated source RF signals are supplied to at least one lower electrode and / or at least one upper electrode.

[0023] The second RF generation unit 31b is electrically connected to or coupled to at least one lower electrode and is configured to generate a bias RF signal (bias RF power). In one embodiment, the second RF generation unit 31b is electrically connected to or coupled to at least one lower electrode via at least one impedance matcher. When the first RF generation unit 31a is electrically connected to or coupled to a lower electrode, the second RF generation unit 31b may be electrically connected to or coupled to the same lower electrode, or it may be electrically connected to or coupled to a different lower electrode. The frequency of the bias RF signal may be the same as or 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 100 kHz to 60 MHz. In one embodiment, the second RF generation unit 31b may 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 various embodiments, at least one of the source RF signal and the bias RF signal may be pulsed.

[0024] The power supply system 30 may also include a power supply 32 that is electrically connected to or coupled to the plasma processing chamber 10. The power supply 32 includes a first voltage generation unit 32a and a second voltage generation unit 32b. In one embodiment, the first voltage generation unit 32a is electrically connected to or coupled to at least one lower electrode and is configured to generate a first voltage signal. The generated first voltage signal is applied to at least one lower electrode. In one embodiment, the second voltage generation unit 32b is electrically connected to or coupled to at least one upper electrode and is configured to generate a second voltage signal. The generated second voltage signal is applied to at least one upper electrode.

[0025] In various embodiments, the first and / or second voltage signals may be pulsed. In this case, the first voltage generation unit 32a and / or the second voltage generation unit 32b function as voltage pulse generation units configured to generate a sequence of voltage pulses. Thus, the sequence of voltage pulses is applied to at least one lower electrode and / or at least one upper electrode. In one embodiment, the sequence of voltage pulses has a plurality of cycles, each cycle including a burst of voltage pulses in a first period and a constant reference voltage in a second period. That is, in the sequence of voltage pulses, the burst of voltage pulses is repeated. The absolute value of the voltage level of the voltage pulse is greater than the absolute value of the voltage level of the reference voltage. The voltage pulse may have an arbitrary waveform having a rectangle, trapezoid, triangle, or a combination thereof, and the arbitrary waveform may change over time. The voltage pulse may have positive polarity or negative polarity. The sequence of voltage pulses may also include one or more positive voltage pulses and one or more negative voltage pulses within one cycle. The first and second voltage generation units 32a and 32b may be provided in addition to the power supply 31, and the first voltage generation unit 32a may be provided in place of the second RF generation unit 31b.

[0026] The exhaust system 40 may be connected to, for example, a gas outlet 10e located at the bottom of the plasma processing chamber 10. The exhaust system 40 may include a pressure regulating valve and a vacuum pump. The pressure regulating valve regulates the pressure in the plasma processing space 10s. The vacuum pump may include a turbomolecular pump, a dry pump, or a combination thereof.

[0027] The plasma processing apparatus 1 is equipped with various measuring instruments not shown in the figure. These measuring instruments include sensors that measure power output values ​​such as voltage, current, and power, values ​​of variable electrical elements such as capacitors and coils in the matching unit, flow rates of various gases used, temperatures of various device components, pressure in the plasma processing chamber 10, opening degree of the pressure control valve, valve open / closed state, and gas exhaust speed at 1 tick intervals. 1 tick represents a sampling interval (10 ms, 100 ms, etc.) determined for each sensor. The sensors measure the object to be measured at 1 tick intervals and output time-series data related to the obtained measured values ​​to the control unit 2.

[0028] The above measuring instruments may include instruments that measure the state of the gas supplied into the plasma processing chamber 10, the state of the plasma generated in the plasma processing chamber 10, and the state of the products generated in the plasma processing chamber 10. These measuring instruments include a mass spectrometer that identifies atoms and molecules emitted from the object to be processed, a plasma emission monitor using optical spectral measurement, and a deposit monitor in the processing chamber using infrared spectroscopy. These measuring instruments measure the object to be measured every tick and output time-series data related to the obtained measurement values ​​to the control unit 2.

[0029] The above measuring instruments may include instruments for measuring the state of the substrate W during plasma processing. These measuring instruments include an optical monitor for imaging the appearance of the substrate W, a dimensional measuring device for measuring the processed shape and film thickness of the substrate W, and a temperature measuring device for measuring the surface temperature of the substrate W. These measuring instruments measure the object to be measured every tick and output time-series data related to the obtained measured values ​​to the control unit 2.

[0030] The following describes the overview of the processing performed by the processing unit 2a1. Figure 3 is an explanatory diagram illustrating the overview of the processing performed by the processing unit 2a1. The control unit 2 controls each element of the plasma processing apparatus 1 according to the plasma processing recipe (hereinafter also simply referred to as the recipe) created by the user, and applies the plasma processing desired by the user to the substrate W to be processed. The recipe consists of multiple steps, and setting values ​​such as control values ​​and target values ​​for each element are defined for each step.

[0031] When plasma processing is performed using the same recipe, it is preferable to obtain identical results within a reasonable margin of error. However, in actual control, due to the complexity of the plasma processing system and discontinuities such as device startup / shutdown, control abnormalities such as hunting, spikes, drops, and bimodal behavior may occur depending on the recipe. When such control abnormalities occur, it becomes difficult to obtain identical results even when using the same recipe.

[0032] Conversely, if a stable recipe can be used for controlling the plasma processing apparatus 1, plasma processing under unstable control can be avoided, and the same results can be obtained within a margin of error. Also, if it can be determined in advance that the recipe for controlling the plasma processing apparatus 1 is not stable, improvements such as changing the set values ​​can be made.

[0033] Therefore, in this embodiment, a method for determining the stability of a recipe is disclosed with respect to the control of the plasma processing apparatus 1. Specifically, the processing unit 2a1 determines the stability of a recipe by performing (1) a score check, (2) a margin check using a prediction formula, and (3) a margin check using a moving average. In the following description, the recipe to be judged (for example, a newly created recipe) will be referred to as the base recipe.

[0034] (1) Score check In the score check, the processing unit 2a1 calculates a score for each step specified by the base recipe from the log data (time-series data obtained from measuring instruments) when the plasma processing is performed according to the base recipe. The processing unit 2a1 determines the stability of the base recipe by comparing the calculated score with a reference value.

[0035] (2) Margin check using prediction formula In the margin check using prediction formula, the processing unit 2a1 creates a prediction formula that predicts the sensor value when the set value specified in the base recipe is changed. The processing unit 2a1 calculates the difference between the predicted value from the prediction formula and the actual sensor value, and determines the stability of the base recipe by comparing the calculated difference with the reference value.

[0036] (3) Margin check using moving average In the margin check using moving average, the processing unit 2a1 calculates the difference between the time series data that was actually measured and the data that has been processed with moving average, and determines the stability of the base recipe based on the calculated difference.

[0037] The processing unit 2a1 may perform any one of the checks (1) to (3) to determine the stability of the base recipe, or it may perform any two of the checks, or all three of the checks, to determine the stability of the base recipe.

[0038] The details of the score check are described below. Figure 4 is an explanatory diagram illustrating the details of the score check. The control unit 2 controls the plasma processing apparatus 1 according to the base recipe and performs plasma processing on multiple substrates W (for example, 25 substrates W). The control unit 2 acquires time-series data obtained from measuring instruments during the execution of plasma processing via the communication interface 2a3 and stores it as log data in the storage unit 2a2.

[0039] The processing unit 2a1 reads log data from the storage unit 2a2 and calculates statistical indicators for the log data for each tick. The processing unit 2a1 calculates the variability between log data (for example, the sample standard deviation using equation 1) as a statistical indicator.

[0040]

[0041] Here, n represents the number of log data (i.e., the number of substrates W that were plasma-treated using the base recipe). SVi is the measured value at the same sampling time for each log data, and the SV bar represents the average value of the measured value SVi.

[0042] Alternatively, the processing unit 2a1 may calculate, as a statistical index, a value obtained by dividing the sample standard deviation by the average value or a value obtained by dividing the sample standard deviation by the median value. When using time-series data from a measuring instrument (sensor) where the magnitude of the absolute value has no meaning, the processing unit 2a1 may calculate the sample standard deviation as the statistical index. When using time-series data from a measuring instrument (sensor) where the magnitude of the absolute value has meaning, the processing unit 2a1 may calculate, as a statistical index, a value obtained by dividing the sample standard deviation by the average value or a value obtained by dividing the sample standard deviation by the median value. As the statistical index, the standard deviation may be used instead of the sample standard deviation.

[0043] Alternatively, the processing unit 2a1 may calculate, as a statistical index, the ratio between the first time-series data and the second time-series data measured as a response to the first time-series data. As such a statistical index, the processing unit 2a1 can calculate, for example, the ratio of the reflected power to the incident power in plasma processing.

[0044] Based on the statistical indexes (sample standard deviation, sample standard deviation / average value, sample standard deviation / median value) calculated for each tick, the processing unit 2a1 calculates the maximum value of the statistical index at each step and determines the calculated maximum value as the score for that step. Instead of calculating the score (the maximum value of the statistical index) for each step, the score may be calculated only for some representative steps, or the maximum value of the statistical index for a plurality of steps may be calculated as the score for that step interval.

[0045] The processing unit 2a1 compares the calculated score with a pre-set reference value and determines the stability of the base recipe based on the comparison result. The reference value is set based on the results of mass production recipes performed in the past. For example, if control abnormalities such as hunting, spikes, or bimodal occurred in a mass production recipe performed in the past, the maximum value of the statistical indicator can be calculated by referring to the log data of the plasma processing in question, and the reference value can be set based on the calculated maximum value. Note that the reference value does not need to be set for each step. For example, each step of a mass production recipe can be classified into one or more main steps representing the main processing and transition steps connecting the main steps, and a reference value can be set for both the main steps and the transition steps.

[0046] The processing unit 2a1 determines that the base recipe (the relevant step in the base recipe) is stable if the calculated score is lower than the reference value, and that the base recipe (the relevant step in the base recipe) is unstable if the calculated score is higher than the reference value. Since the processing unit 2a1 can determine the stability of each step in the base recipe, it becomes possible to prioritize the steps that need improvement.

[0047] Next, the details of the margin check using the prediction formula will be explained. The processing unit 2a1 creates a prediction formula that predicts the sensor value when the setpoints specified in the base recipe are varied. Figure 5 is an explanatory diagram illustrating how the setpoints are varied. The processing unit 2a1 increases or decreases the setpoints specified in the base recipe to create several conditions. For example, as shown in Figure 5, the processing unit 2a1 determines three levels of setpoints (0%, ±10%) for pressure by increasing or decreasing them by ±10% from the setpoint in the base recipe, and including the case where the setpoint is not varied. Similarly, the processing unit 2a1 determines five levels of setpoints (0%, ±5%, ±10%) for source RF power (HF) and bias RF power (LF) by increasing or decreasing them by ±5% and ±10%, and including the case where the setpoint is not varied. From these combinations of setpoints, the processing unit 2a1 can create a total of 27 conditions. In the example in Figure 5, the three overlapping conditions (shown by knitting in the figure) have been removed, taking into consideration that when LF is BSL (0%), HF is also BSL (0%).

[0048] The way of setting values is not limited to the example of FIG. 5, and the setting values may be appropriately increased or decreased so that three or more levels of setting values can be obtained for each setting value. In the example of FIG. 5, three types of setting values are increased or decreased, but the number of setting values to be increased or decreased is not limited to three types, and may be four or more types.

[0049] The control unit 2 controls the plasma processing apparatus 1 according to each of a plurality of conditions created by setting the setting values of the base recipe, and performs plasma processing on a plurality of substrates W (for example, three substrates W) under each condition. During the execution of the plasma processing, the control unit 2 acquires time-series data obtained from various measuring instruments through the communication interface 2a3, and stores it in the storage unit 2a2 as log data.

[0050] FIG. 6 is a graph showing an example of log data. The horizontal axis of the graph represents time, and the vertical axis represents the measured value (Upper C1 Position) related to the capacitance of the variable capacitor in the matcher. The graph of FIG. 6 shows the time change of the measured value related to the capacitance of the variable capacitor when substrate processing is performed on three substrates for each of the above 27 conditions. It can be seen from the graph that by changing the setting value of the recipe, the measured value fluctuates. Although FIG. 6 shows the fluctuation of the measured value related to the capacitance of the variable capacitor, when the setting value of the recipe is changed, not only the measured value related to the capacitance, but also the measured values indicated by various measuring instruments fluctuate.

[0051] [[ID=ID=10]]The processing unit 2a1 reads the log data from the storage unit 2a2, and creates a prediction formula based on the read log data. Since various measured values (sensor outputs) when the setting values of the base recipe are set are recorded in the log data, a prediction formula for predicting the measured value from the setting value can be created by using the log data.

[0052] For example, when the setting value of the pressure is press [%], the setting value of the HF is HF [%], and the setting value of the LF is LF [%], the prediction formula of the measured value (sensor output) is described as Equation 2.

[0053]

[0054] Here, SV is the predicted value of the measured value (sensor output), a, b, and c are the coefficients of each term, and offset is the offset.

[0055] Plasma processing is performed under various conditions created by varying the set values ​​of pressure, HF, and LF. When the measured values ​​obtained under each condition are found as log data, the processing unit 2a1 can determine the coefficients a, b, c, and offset using existing methods such as the Partial Least Squares (PLS) method.

[0056] Once the coefficients a, b, c and offset are determined, the processing unit 2a1 can use the obtained prediction formula to predict the measured values ​​for any set values ​​of pressure, HF, and LF. Note that the values ​​to be predicted are not limited to the measured values ​​of pressure, HF, and LF. For example, the prediction formula can predict measured values ​​measured by various measuring instruments, such as values ​​related to the capacitance of a variable capacitor (Upper C1 Position, Lower C1 Position, etc.) and the frequency value of the RF signal generated by the AC plasma generation unit. Furthermore, multiple prediction formulas may be prepared for each type of measuring instrument. In this case, the coefficients a, b, c, and offset should be determined for each type of measuring instrument, and a prediction formula should be derived for each.

[0057] Figure 7 is a graph showing a comparison of measured values ​​and predicted values. The upper graph in Figure 7 shows a comparison between the actual measured value of the capacitance value (Upper C1 Position) of the variable capacitor in a specific step interval (for example, the interval between steps 20 and 29 described later) and the predicted value obtained by the prediction formula. When board processing is performed on a set number of boards for each condition while changing the conditions (settings for pressure, HF, and LF), measured values ​​are obtained for the number of conditions × the number of boards set for each condition. For example, if board processing is performed on 3 boards for each of the 27 conditions above, measured values ​​for a total of 81 boards are obtained. In the upper graph of Figure 7, the measured values ​​for each of the 81 boards processed are shown as solid lines, and the predicted values ​​for the corresponding conditions (settings) are shown as dashed lines.

[0058] The lower graph in Figure 7 shows the difference between the measured value and the predicted value. This example shows that when the 46th cumulative substrate was processed under conditions where the pressure was reduced by 10% and the LF was increased by 5% from the base recipe settings, an abnormality occurred in the measured value. Note that if bimodal (two-peak or binarization) occurs in the measured value of the substrate processing, the absolute value of the difference between the measured value and the predicted value will be large, so the abnormality in the measured value is thought to be due to a control abnormality accompanied by bimodal.

[0059] The processing unit 2a1 calculates the difference between the actual measured value and the predicted value obtained by the prediction formula, and compares the absolute value of this difference with the set reference value. Figure 8 is a chart showing a comparison example with the reference value. The example in Figure 8 shows the results of calculating the difference between the measured value and the predicted value for each of the 27 conditions above for each specific step interval and comparing it with the set reference value. The reference value may be determined artificially by the user, or the processing unit 2a1 may determine an appropriate value (for example, the median or mean of the measured values).

[0060] Figure 8 shows the number of substrates whose calculated difference exceeded the standard value, indicated by numbers from 0 to 3. For example, in the step section from step 20 to 29, under conditions where the pressure was reduced by 10% and the LF was increased by 5% from the base recipe settings, one substrate exceeded the standard value, while under other conditions, no substrates exceeded the standard value. The same applies to the other step sections, and it can be seen that multiple substrates exceeded the standard value in the step sections of step 30 and from step 13 to 18.

[0061] If the absolute value of the calculated difference exceeds the reference value, the processing unit 2a1 determines that the recipe has low stability because the margin relative to the set value is small. Conversely, if the absolute value of the calculated difference is smaller than the reference value, the processing unit 2a1 determines that the recipe has high stability because the margin relative to the set value is large.

[0062] Next, we will explain margin checking using moving averages. Figure 9 is a graph showing another example of log data. The horizontal axis of the graph represents time, and the vertical axis represents the measured value of the magnetic field near the upper electrode (Upper Mag.). The graph shows the time change of the magnetic field near the upper electrode when three substrates were treated for each of the 27 conditions mentioned above. From the graph, it can be seen that changing the recipe settings causes a change in the magnitude of the magnetic field. Figure 9 shows the change in the magnitude of the magnetic field, but when the recipe settings are changed, changes occur not only in the magnitude of the magnetic field, but also in the measured values ​​shown by various measuring instruments.

[0063] The graph in Figure 9 shows an example where the variation between measured values ​​is relatively large in steps 12 and 30, and relatively small in step 39, although hunting exists within the step. If the score check and margin check using the prediction formula described above are applied, the recipe may be judged as unstable in steps 12 and 30, and as stable in step 39.

[0064] In this embodiment, in order to detect recipe instability that is difficult to detect by score checks or margin checks using prediction formulas, the processing unit 2a1 performs a margin check using a moving average. In a margin check using a moving average, the processing unit 2a1 calculates a moving average of the measured values, takes the difference between the actual measured values ​​and the calculated moving average, and determines the stability of the recipe by comparing the calculated difference with a reference value.

[0065] Figure 10 is a graph showing an example of calculating the difference between a measured value and a moving average. In each graph, the horizontal axis represents time, and the vertical axis represents the difference between the measured value and the moving average. The processing unit 2a1 calculates the moving average from the measured values ​​for each condition. The time interval for calculating the moving average is set as appropriate. For each condition, the processing unit 2a1 calculates the difference between the measured value and the moving average every tick. The difference to be calculated may simply be measured value - moving average, or it may be ((measured value - moving average) / moving average) × 100.

[0066] The graphs in the upper, middle, and lower sections of Figure 10 show the calculated difference values ​​in steps 12, 30, and 39, respectively. From the graphs in Figure 10, it can be seen that the calculated difference values ​​are relatively small in steps 12 and 30, and relatively large in step 39.

[0067] The processing unit 2a1 can determine recipe instability caused by control anomalies accompanied by hunting by appropriately setting a reference value for the difference between the measured value and the moving average and comparing the calculated difference with the reference value. Alternatively, the instability of the recipe caused by control anomalies accompanied by spikes may be determined by appropriately setting a reference value for the maximum value of the difference between the measured value and the moving average and comparing the maximum value of the difference with the reference value.

[0068] Next, the procedure for processing performed by the processing unit 2a1 will be described. Figure 11 is a flowchart illustrating the procedure for processing performed by the processing unit 2a1. The processing unit 2a1 acquires a base recipe (step S101). If a base recipe created by the user exists on the user's terminal, the processing unit 2a1 can acquire the base recipe by communicating with the user's terminal via the communication interface 2a3. The base recipe consists of multiple steps, and setting values ​​such as control values ​​and target values ​​for each element are defined at the step level. The processing unit 2a1 stores the acquired recipe in the storage unit 2a2.

[0069] The processing unit 2a1 performs substrate processing according to the acquired base recipe (step S102). The processing unit 2a1 controls each element of the plasma processing apparatus 1 according to the setting values ​​for each step described in the base recipe, and performs substrate processing on, for example, 25 substrates in one lot.

[0070] The processing unit 2a1 acquires time-series data of measured values ​​measured in a time-series manner by various measuring instruments during the execution of substrate processing (step S103). The processing unit 2a1 stores the acquired time-series data as log data in the storage unit 2a2. When substrate processing is performed on 25 substrates in one lot, log data for 25 substrates is obtained.

[0071] The processing unit 2a1 calculates the score of each measuring instrument for each step of the recipe (step S104). For example, the processing unit 2a1 calculates the variability between log data (sample standard deviation) for each tick as a statistical indicator of the log data. Alternatively, the processing unit 2a1 may calculate the value obtained by dividing the sample standard deviation by the mean, or the value obtained by dividing the sample standard deviation by the median, as a statistical indicator. The processing unit 2a1 may also calculate the ratio of the first time series data to the second time series data measured as a response to the first time series data as a statistical indicator. For example, the processing unit 2a1 can calculate the ratio of reflected power to incident power in plasma processing.

[0072] The processing unit 2a1 calculates the maximum value of the statistical indicator at each step based on the statistical indicator calculated for each tick, and sets the calculated maximum value as the score for that step.

[0073] The processing unit 2a1 compares the calculated score with a first reference value to determine the stability of the base recipe (step S105). Here, the first reference value is a reference value set for the score, and is set based on the results of mass production recipes implemented in the past. If the score exceeds the first reference value, the processing unit 2a1 determines that the base recipe is not stable, and if the score does not exceed the first reference value, it determines that the base recipe is stable. If the processing unit 2a1 determines that the base recipe is not stable, it may change the setting value in the base recipe and return to step S102, and determine the stability of the recipe based on the changed setting value.

[0074] After determining stability through score checking, the processing unit 2a1 changes the base recipe settings and creates multiple conditions in order to perform a margin check using a prediction formula (step S106).

[0075] The processing unit 2a1 performs substrate processing according to each condition (step S107). The processing unit 2a1 controls each element of the plasma processing apparatus 1 according to each condition modified from the base recipe, and performs substrate processing on, for example, three substrates per condition.

[0076] The processing unit 2a1 acquires time-series data of measured values ​​measured in a time-series manner by various measuring instruments during the execution of substrate processing (step S108). The processing unit 2a1 stores the acquired time-series data as log data in the storage unit 2a2. If the number of conditions created in step S106 is 27, and three substrates are processed per condition, then log data for 81 substrates will be obtained.

[0077] The processing unit 2a1 reads log data from the storage unit 2a2 and creates a prediction formula to predict sensor values ​​from the set values ​​(step S109). Since the measured values ​​obtained as log data when the set values ​​are varied from the base recipe, the processing unit 2a1 can create a prediction formula by using a known method such as the partial least squares method.

[0078] The processing unit 2a1 calculates the difference between the actual measured value and the predicted value predicted from the prediction formula, and determines the stability of the recipe by comparing the calculated difference with a second reference value (step S110). Here, the second reference value is a reference value set for the difference between the measured value and the predicted value, and is set as appropriate by the user or the processing unit 2a1.

[0079] If the difference exceeds the second reference value, the processing unit 2a1 determines that the base recipe is unstable, and if the difference does not exceed the first reference value, it determines that the base recipe is stable. If the processing unit 2a1 determines that the base recipe is unstable, it may change the settings in the base recipe and return to step S102, and perform the score check and margin check using the prediction formula again.

[0080] After completing the margin check using the prediction formula, the processing unit 2a1 performs a margin check using a moving average.

[0081] The processing unit 2a1 reads log data from the storage unit 2a2 and calculates a moving average for each measured value of the log data (step S111). The processing unit 2a1 calculates the difference between the actual measured value and the moving average and determines the stability of the recipe by comparing the calculated difference with a third reference value (step S112). Here, the third reference value is a reference value set for the difference between the measured value and the moving average, and is set as appropriate by the user or the processing unit 2a1.

[0082] If the difference exceeds the third reference value, the processing unit 2a1 determines that the base recipe is unstable, and if the difference does not exceed the third reference value, it determines that the base recipe is stable. If the processing unit 2a1 determines that the base recipe is unstable, it may change the settings in the base recipe and return to step S102, and perform the score check, margin check using the prediction formula, and margin check using the moving average again.

[0083] In the flowchart in Figure 11, the procedure involves performing score checks, margin checks using prediction formulas, and margin checks using moving averages. However, it is also acceptable to perform only one of these steps to determine the stability of the recipe, or to combine any two of them to determine the stability of the recipe.

[0084] The flowchart in Figure 11 shows a procedure in which score checking, margin checking using a prediction formula, and margin checking using a moving average are performed in this order. However, the order of these executions can be changed as appropriate, or they can be performed simultaneously.

[0085] The processing unit 2a1 may control the plasma processing apparatus 1 according to a recipe that it has determined to be stable, and perform substrate processing for each lot.

[0086] The processing unit 2a1 may present the comparison results obtained during each check process to the user. Figure 12 is a schematic diagram showing a first example of presentation. Figure 12 shows an example in which the comparison result between the score calculated for the measured value related to the capacitance of the variable capacitor (Upper C1 Position) and the reference value (first reference value) is displayed as a bar graph during the score check. Such a graph can be displayed on the user's terminal, for example. Figure 12 shows an example in which, if the score does not meet the reference value, the base recipe settings are changed and the score check is performed multiple times. The latest base recipe (New BKM) meets the reference value in all step intervals indicated by 1 to 9, indicating that a highly stable recipe has been obtained.

[0087] Figure 13 is a schematic diagram illustrating a second example. Figure 13 shows an example where, during a margin check using a prediction formula, the difference between the measured value and the predicted value, and the comparison result with a reference value (second reference value), are displayed on the screen. Such a screen is displayed, for example, on the user's terminal. In Figure 13, the number of substrates whose calculated difference exceeded the reference value is shown by numbers from 0 to 3, and the conditions under which the number of substrates exceeded the reference value are highlighted in color. For example, in the step section from step 20 to 29, it is shown that under the conditions where the pressure was reduced by -10% and the LF was increased by 5% from the base recipe settings, there was one substrate that exceeded the reference value, while under the other conditions, there were no substrates that exceeded the reference value. The same applies to other step sections, and it can be seen that there are multiple substrates that exceeded the reference value in step 30 and the step section from step 13 to 18. By viewing such a screen, the user can easily identify the steps in the base recipe that need improvement.

[0088] The embodiments disclosed herein should be considered in all respects to be illustrative and not restrictive. The scope of the invention is indicated by the claims, not in the sense described above, and all modifications within the sense and scope equivalent to the claims are intended.

[0089] In the embodiment, an example of application to a capacitively coupled plasma processing apparatus 1 was described, but it is not limited to capacitive coupling; it can be applied to any type of plasma processing apparatus, including Inductively Coupled Plasma (ICP), Radial Line Slot Antenna (RLSA), Electron Cyclotron Resonance Plasma (ECR), and Helicon Wave Plasma (HWP).

[0090] In this embodiment, wafer W was used as an example of a substrate to be processed. However, the substrate to be processed is not limited to wafer W, but may be various substrates used in FPDs (Flat Panel Displays), printed circuit boards, etc.

[0091] 1 Plasma processing apparatus 2 Control unit 2a Computer 2a1 Processing unit 2a2 Memory unit 2a3 Communication interface 10 Plasma processing chamber 11 Substrate support unit 20 Gas supply unit 30 Power supply

Claims

1. An information processing method that uses a computer to perform the following steps: acquire time-series data measured by a specific sensor when plasma processing is performed for each of several conditions obtained by changing one or more set values ​​in a plasma processing recipe; calculate the average value of the time-series data in a specific step interval of the plasma processing recipe for each of the several conditions; derive a correlation between the set value and the calculated average value; compare the calculated average value with a predicted value predicted from the correlation for each of the several conditions; and determine the stability of the plasma processing recipe based on the comparison result between the average value and the predicted value.

2. The information processing method according to claim 1, wherein the computer performs a process to calculate a statistical index relating to the acquired time-series data and to determine the stability of the plasma processing recipe based on the calculated statistical index.

3. The information processing method according to claim 1, wherein the computer performs a process of calculating the maximum value of the statistical index in a specific step interval in the plasma processing recipe, comparing the calculated maximum value with a reference value for the maximum value, and determining the stability of the plasma processing recipe based on the comparison result between the maximum value and the reference value.

4. The information processing method according to claim 3, wherein the maximum value of a statistical index calculated with respect to plasma treatment performed in the past is set as the reference value.

5. The information processing method according to claim 2, wherein the statistical indicator is the sample standard deviation of the time series data, the value obtained by dividing the sample standard deviation by the mean, the value obtained by dividing the sample standard deviation by the median, or the ratio of the first time series data to the second time series data measured as a response to the first time series data.

6. The information processing method according to claim 5, wherein the ratio is the ratio of reflected power to incident power in the plasma processing.

7. An information processing method according to any one of claims 1 to 6, wherein the computer performs a process to determine the stability of the plasma processing recipe based on the comparison result between the moving average value and the time series data, for each of the multiple conditions, calculates a moving average value of the acquired time series data, compares the calculated moving average value with the acquired time series data, and 8. An information processing method that, for each of the multiple conditions obtained by changing one or more setting values ​​in a plasma processing recipe, acquires time-series data measured by a specific sensor when plasma processing is performed, calculates a moving average of the acquired time-series data for each of the multiple conditions, compares the calculated moving average with the acquired time-series data, and performs a process by computer to determine the stability of the plasma processing recipe based on the comparison result between the moving average and the time-series data.

9. The information processing method according to claim 8, wherein the computer performs a process to calculate a statistical index relating to the acquired time-series data and to determine the stability of the plasma processing recipe based on the calculated statistical index.

10. An information processing device comprising at least one processor, the processor acquires time-series data measured by a specific sensor when plasma processing is performed for each of a plurality of conditions obtained by changing one or more set values ​​in a plasma processing recipe, calculates the average value of the time-series data in a specific step interval of the plasma processing recipe for each of the plurality of conditions, derives a correlation between the set value and the calculated average value, compares the calculated average value with a predicted value predicted from the correlation for each of the plurality of conditions, and determines the stability of the plasma processing recipe based on the comparison result between the average value and the predicted value.

11. An information processing device comprising at least one processor, wherein the processor acquires time-series data measured by a specific sensor when plasma processing is performed for each of a plurality of conditions obtained by changing one or more setting values ​​in a plasma processing recipe, calculates a moving average value of the acquired time-series data for each of the plurality of conditions, compares the calculated moving average value with the acquired time-series data, and determines the stability of the plasma processing recipe based on the comparison result between the moving average value and the time-series data.

12. A computer program that causes a computer to perform the following processes: acquire time-series data measured by a specific sensor when plasma processing is performed for each of several conditions obtained by changing one or more setting values ​​in a plasma processing recipe; calculate the average value of the time-series data in a specific step interval of the plasma processing recipe for each of the several conditions; derive a correlation between the setting value and the calculated average value; compare the calculated average value with a predicted value predicted from the correlation for each of the several conditions; and determine the stability of the plasma processing recipe based on the comparison result between the average value and the predicted value.

13. A computer program that causes a computer to perform a process to determine the stability of the plasma processing recipe, for each of several conditions obtained by changing one or more setting values ​​in the plasma processing recipe, by acquiring time-series data measured by a specific sensor when plasma processing is performed, by calculating a moving average of the acquired time-series data for each of the several conditions, by comparing the calculated moving average with the acquired time-series data, and by determining the stability of the plasma processing recipe based on the comparison result between the moving average and the time-series data.

Images