Dynamic Process Control in Semiconductor Manufacturing

Abstract

Methods and systems are provided for dynamic process control of substrate processing, for example, in semiconductor manufacturing applications. Some exemplary systems and methods are provided for advanced monitoring and machine learning in atomic layer deposition (ALD) processes. Some examples also relate to dynamic process control and monitoring for chamber parameter matching and gas line fill times.

Description

Technical Field

[0001] [Claims of Priority] This patent application claims the benefit of priority of U.S. Provisional Patent Application No. 62 / 885,667, entitled "Dynamic Process Control In Semiconductor Manufacturing," filed on August 12, 2019 by Kumar et al., which is hereby incorporated by reference in its entirety.

[0002] The present disclosure generally relates to dynamic process control in substrate processing, and in some examples, to systems and methods for advanced monitoring and machine learning in atomic layer deposition (ALD) processes. Some examples are also related to dynamic process control and monitoring for chamber matching, and filling times of gas lines.

Background Art

[0003] Currently, most, if not all, of the parameters associated with a substrate processing chamber are monitored to operate near the set values of the components. For example, the flow of a mass flow controller (MFC) and the chamber pressure may include a certain error margin. To account for this error, it is common to make the upper or lower limits of the parameter settable to a certain value or a certain ratio. For example, in an ALD process, the opening and closing times of valves are monitored and the times are appropriately reflected in the monitoring parameters.

[0004] However, current process monitoring methods are generally only suitable for detecting relatively broad or general malfunctions of the processing chamber or its components. Such general detection is not a problem in steady-state or single-step situations, such as chemical vapor deposition (CVD) or plasma-enhanced chemical vapor deposition (PECVD) processes, but has limitations for use or application in multi-step processes such as ALD where the chamber state changes in milliseconds.

[0005] The background art provided herein is for the purpose of generally presenting the context of the present disclosure. Within the scope described in this background art section, research by the inventors named at the present time, as well as aspects of the description that cannot be separately regarded as prior art at the time of filing, are not recognized as prior art against the present disclosure, whether explicitly or implicitly.

Summary of the Invention

[0006] The present disclosure generally relates to dynamic process control in a semiconductor substrate manufacturing system. In some examples, systems and methods for advanced monitoring and machine learning in an ALD process are provided. Some examples also relate to dynamic process control and monitoring for chamber alignment, and the fill time of gas lines. Some examples are described in the context of semiconductor processing, but are equally applicable to substrate processing outside of such a context, such as the processing of dielectrics such as metals and photomasks.

[0007] In some examples, consecutive ALD cycles are monitored individually, and all steps of the ALD cycle are matched using curve fitting or a defined error margin. All ALD cycles are repeatable in some examples, and a time reference can be defined to monitor the repeatability of different variables. In some examples, a comparison of consecutive ALD cycles for different measured variables may be performed. Examples of variables may include chamber pressure, precursor supply pressure (or precursor manifold pressure), reflected radio frequency (RF) energy, forward power, and burst purge pressure, among others. Also, other monitoring variables and parameters may be used.

[0008] Some examples provide tool warnings or error messages based on deviations from monitored values. Some examples include algorithms or software for achieving these goals.

[0009] An example of an embodiment provides a system for monitoring a processing cycle in an atomic layer deposition (ALD) semiconductor manufacturing process. An example of the system includes a processing chamber for an ALD manufacturing process and one or more controllers configured to perform process monitoring operations, the operations including defining a time reference for an ALD cycle based on repetitive operations in the manufacturing process, accessing a golden curve including a series of parameter values for a series of data points at periodic time increments based on the time reference, accessing variability or tolerance margins for each data point in the golden curve, collecting parameter data based on a periodic time increment of one cycle in the ALD manufacturing process, dynamically monitoring whether a parameter value of the parameter data is within the variability or tolerance margin at a certain data point, and adjusting the manufacturing process such that a parameter value in a subsequent cycle matches a related parameter value in the golden curve based on a determination that the parameter value has deviated from the variability or tolerance margin.

[0010] In some examples, the repetitive operations forming the basis of the time reference include opening and closing a specific valve that supplies the processing chamber. In some examples, the parameter data is collected at regular intervals based on a collection frequency, and the collection frequency is in the range of 0 to 1 Hz, 1 to 10 Hz, 10 to 100 Hz, or 100 to 1000 Hz.

[0011] In some examples, the regular intervals are based on trigger points in the ALD manufacturing process, and each trigger point defines or is based on a point in time in a step within the ALD manufacturing process.

[0012] In some examples, the operations further include comparing parameter data collected at a trigger point with a corresponding parameter data set in the golden curve. In some examples, the parameter data includes parameter values related to one or more of a precursor manifold pressure, a purge pressure, a conversion manifold pressure, a chamber pressure, a gas flow, an RF reflected power, and an RF forward power.

[0013] In a further example, an example of a system includes a processing chamber for a manufacturing process and one or more controllers configured to perform a process monitoring operation, the operation including identifying parameters of the manufacturing process, generating a first curve of parameter values including a first parameter value based on a first cycle of the manufacturing process, identifying a second value of the parameter for a second cycle of the manufacturing process based on the curve generated for the first cycle, and adjusting the manufacturing process to match the first parameter value with the second parameter value.

[0014] In some examples, the operation further includes generating a second curve including a plurality of parameter values derived from a second cycle of the manufacturing process and curve fitting the second curve of parameter values to the first curve of parameter values.

[0015] In some examples, the curve fitting operation includes fitting a series of parameter value curves based on or derived from the third and subsequent cycles of the manufacturing process to the first or second curve to generate a golden curve defining a set of golden parameter values for the manufacturing process. In some examples, each cycle of the manufacturing process includes a plurality of steps of an ALD process, and the operation further includes matching the parameter values at each step of the ALD process with the parameter values in the set of golden parameter values.

[0016] In some examples, the golden curve includes golden parameter values for each step of the ALD process. In some examples, the identified parameters are associated with control variables of the manufacturing process, and the operation further includes identifying a matching value between the value of the control variable in the second cycle and the value of the control variable in the first cycle using the first and second parameter values directly or indirectly, and adjusting the manufacturing process to match the first parameter value with the second parameter value.

[0017] In a further example, a self - learning system is provided for monitoring processing steps in a semiconductor manufacturing cycle. An example of the system includes a processing chamber that houses a semiconductor manufacturing process including a series of iterative semiconductor manufacturing cycles, each cycle including a plurality of processing steps and one or more controllers associated with the processing chamber configured to perform a process monitoring operation, the operation including generating a set of reference golden parameter values for each step in a series of iterative cycles based on parameter data collected from the processing chamber, generating a machine learning model based on the set of reference golden values, and using the machine learning model to match parameter values in a second cycle in a series of iterative cycles with corresponding parameter values in a first cycle in the series of iterative cycles.

[0018] In some examples, the semiconductor manufacturing process is an ALD process, and the steps of each cycle in a series of iterative cycles include a dose step, a purge step, a conversion step, and continuous steps including a purge step.

[0019] In some examples, the operation further includes generating a golden curve including parameter value data at each continuous step and using the golden curve as training data for the machine learning model.

[0020] In some examples, the operation further includes iterating and matching each cycle in a series of iterative cycles by matching parameter values in each continuous step of a second cycle with corresponding parameter values in each continuous step of a first cycle based on the machine learning model.

[0021] In some examples, the parameter includes one or more of precursor manifold pressure, purge pressure, chamber pressure, gas flow, chamber temperature, RF reflected power, and RF forward power.

[0022] In some examples, the operation further includes collecting performance data from a processing chamber, identifying a drift in the performance data, generating performance drift data, and incorporating the drift data into training data for a machine learning model.

[0023] In a further example, a system for monitoring a processing cycle in a semiconductor manufacturing process is provided. Here, an example of the system includes a processing chamber for a manufacturing process and one or more controllers configured to perform process monitoring operations, and the operations include identifying parameters of the manufacturing process and generating a first curve of parameter values including a first parameter value based on a first cycle of the manufacturing process, where the parameter values include a gas line fill time of a line supplying the processing chamber.

[0024] In some examples, the operation further includes calculating the gas line fill time based on a pressure ramp-up time between opening a valve in a line supplying the processing chamber and establishing a subsequent constant pressure increase. In some examples, the operation includes identifying a second value of a parameter for a second cycle of the manufacturing process based on a curve generated for the first cycle, generating a second curve including a plurality of parameter values derived from the second cycle of the manufacturing process, and further curve fitting the second curve of parameter values to the first curve of parameter values.

[0025] In some examples, the curve fitting operation includes fitting a series of parameter value curves based on or derived from the third and subsequent cycles of the manufacturing process to the first or second curve to generate a golden curve defining a set of golden parameter values for the manufacturing process. In some examples, each cycle of the manufacturing process includes a plurality of steps of an ALD process, and the operation further includes matching parameter values at each step of the ALD process to parameter values in the set of golden parameter values.

[0026] In some examples, the golden curve includes golden parameter values for each step of the ALD process.

[0027] In some examples, the identified parameters are associated with control variables of the manufacturing process, and the operation further includes identifying a matching value between the value of the control variable in the second cycle and the value of the control variable in the first cycle using the first and second parameter values directly or indirectly, and adjusting the manufacturing process to match the first parameter value with the second parameter value.

Brief Description of the Drawings

[0028] Some embodiments will be described with reference to the accompanying drawings. These embodiments are illustrative and not shown as limiting.

[0029]

Figure 1

[0030]

Figure 2A

Figure 2B

Figure 2C

[0031]

Figure 3

[0032]

Figure 4

[0033]

Figure 5

[0034]

Figure 6

[0035]

Figure 7

[0036]

Figure 8

[0037]

Figure 9

[0038]

Figure 10

[0039]

Figure 11

[0040]

Figure 12

[0041]

Figure 13

[0042]

Figure 14

[0043]

Figure 15

[0044]

Figure 16

[0045]

Figure 17

[0046]

Figure 18

[0047]

Figure 19

[0048]

Figure 20

[0049]

Figure 21

[0050]

Figure 22

Figure 23

Figure 24

Figure 25

[0051]

Figure 26

DETAILED DESCRIPTION OF THE INVENTION

[0052] The following description includes systems, methods, techniques, instruction sequences, and computing machine program products for implementing exemplary embodiments of the present disclosure. In the following description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the exemplary embodiments. However, it will be apparent to one of ordinary skill in the art that the present disclosure may be practiced without these specific details.

[0053] Part of the disclosure of this patent document may contain materials subject to copyright protection. The copyright owner reserves all copyrights except that, as long as the patent document or patent disclosure is in the patent file wrapper or records of the Patent and Trademark Office, the copyright owner does not object to the patent document or patent disclosure being reproduced by anyone. The following notice applies to all data that constitutes a part of this document, whether described or illustrated hereinafter. Copyright Lam Research Corporation, 2019 - 2020, All Rights Reserved.

[0054] Referring now to FIG. 1, an example of a plasma-based processing chamber is shown. The present subject matter can be used for various semiconductor manufacturing and substrate processing operations, but in the example shown herein, the plasma-based processing chamber will be described assuming plasma-excited or radical-excited CVD or ALD operations. Also, other types of ALD processing techniques (e.g., ALD operations utilizing heat) are known, and it will be understood by those skilled in the art that non-plasma-based processing chambers can also be incorporated. An ALD tool is a special CVD processing system that causes an ALD reaction between two or more chemical species. The two or more chemical species are called precursor gases and are used to form a thin film deposition of a material on a substrate such as a silicon substrate used in the semiconductor industry. The precursor gases are sequentially introduced into the ALD processing chamber and reacted with the substrate surface to form a deposition layer. Generally, the substrate repeatedly interacts with the precursor to gradually deposit one or more material films on the substrate, and the film gradually becomes thicker. In certain applications, various types of films can be formed using multiple precursor gases in the substrate manufacturing process.

[0055] A plasma-based processing chamber 101 is shown in FIG. 1, and this processing chamber 101 is provided with a showerhead 103 (which may also serve as a showerhead electrode) and a substrate support assembly 107. The substrate support assembly 107 may include a pedestal as will be described in detail below. Typically, the substrate support assembly 107 is intended to provide a surface that is substantially isothermal and can function as both a heating element and a heat sink for the substrate 105. The substrate support assembly 107 may include an electrostatic chuck (ESC) that includes a heating element for assisting in the processing of the substrate 105 as described above. The substrate 105 may include a substrate containing an elemental semiconductor (e.g., silicon or germanium), a substrate containing a compound device (e.g., gallium arsenide (GaAs) or gallium nitride (GaN)), or various other types of substrates (including conductor substrates, semiconductor substrates, and non-conductor substrates).

[0056] During operation, the substrate 105 is loaded into the substrate support assembly 107 from the load port 109. The gas line 113 can supply one or more process gases (e.g., precursor gases) to the showerhead 103. In response, the showerhead 103 supplies one or more process gases to the plasma-based processing chamber 101. A gas source 111 (e.g., one or more precursor gas ampoules) for supplying one or more process gases is coupled to the gas line 113. In some examples, an RF power source 115 is coupled to the showerhead 103. In other examples, a power source is coupled to the substrate support assembly 107 or the ESC.

[0057] Downstream of the gas line 113, prior to the inlet of the showerhead 103, a point-of-use (POU) and manifold combination (not shown) controls the entry of one or more process gases into the plasma-based processing chamber 101. In the case of the plasma-based processing chamber 101 used to deposit a thin film in a plasma-enhanced ALD (PEALD) operation, the precursor gases may be mixed within the showerhead 103.

[0058] During operation, the plasma processing chamber 101 is evacuated by a vacuum pump 117. RF power is capacitively coupled between the showerhead 103 and a lower electrode (not shown) provided within or on the substrate support assembly 107. The substrate support assembly 107 is typically supplied with two or more RF frequencies. For example, in various embodiments, the RF frequencies can be selected from at least one of about 1 MHz, 2 MHz, 13.56 MHz, 27 MHz, 60 MHz, and other desired frequencies. It is also possible to provide a coil designed to block or partially block a specific RF frequency, if necessary. Therefore, the specific frequencies described herein are provided merely for ease of understanding. The RF power energizes one or more process gases to form a plasma in the space between the substrate 105 and the showerhead 103. The generated plasma assists in the deposition of various layers (not shown) on the substrate 105. In other applications, the plasma can be used to etch device characteristics into various layers on the substrate 105. The RF power is coupled at least via the substrate support assembly 107. A heater (not shown in FIG. 1) may be incorporated within the substrate support assembly 107. The details of the plasma processing chamber 101 may have various designs.

[0059] In some examples, over the course of a given cycle, ALD can be considered a process of multiple steps (e.g., mainly four steps) including dose, purge, conversion, and purge steps that occur within the processing chamber. Other cycles and steps are also applicable. Unlike a specific PECVD process where parameters such as gas flow, chamber pressure, and RF value are constant throughout the deposition process (cycle), in ALD, these process parameters (etc.) can vary at each step, throughout the cycle, or within consecutive cycles.

[0060] Unmonitored changes in parameters at different steps may obscure variations in some important control variables. For example, chamber pressure has conventionally been required to be maintained at a constant value or a set pressure. Conventionally, variations in gas flow to the chamber at various steps in a multi-step process have been controlled by a throttle valve that moves continuously to maintain the set pressure. However, in changing the gas flow and operating the throttle valve, there is a problem of inherent feedback delay, which can thereby cause poor control, for example, often during an ALD process.

[0061] Similarly, during processing, the precursor manifold pressure in the gas line supplying a precursor gas, such as argon (Ar), often varies due to an increase or a delayed change in gas flow when the precursor gas enters the chamber or is bypassed. In other words, the pressure in a given gas line can vary within a certain range depending on whether the gas is flowing into the chamber (e.g., whether a burst purge pressure is occurring) or is being diverted (e.g., whether a flow pressure limit is occurring). These variations are evident, if not all, in many ALD cycles and are considered undesirable. Deviations from the set values of the parameters may indicate that there is a problem with the processing of the chamber and can affect the characteristics of the substrate or film.

[0062] Referring to FIGS. 2A - 2C, a conventional method of monitoring chamber parameters may include (by way of example only) monitoring chamber pressure (FIG. 2A), precursor manifold pressure (FIG. 2B), and purge pressure (FIG. 2C) around an error band 204 set between a setpoint 202, an upper limit 206, and a lower limit 208. The error band 204 is often relatively large and, in fact, is large enough that smaller but potentially significant variations that occur during the ALD cycle cannot be detected. Therefore, with conventional approaches based on setpoints and error bands, more detailed aspects or chamber states during the ALD cycle cannot be monitored accurately or faithfully. These days, more accurate, controllable, and specific chamber control is required by substrate manufacturers to create high aspect ratio nanometer - sized formations and semiconductor devices on substrates.

[0063] In this regard, refer now to FIG. 3. In some examples, successive dose, purge, conversion, and (again) purge steps within an overall deposition or etch cycle are monitored and coordinated using curve fitting. Some examples include defined error margins configured for each specific step, as opposed to a general error band defined for all cycles of the type described above. In some examples, each step of the ALD cycle (e.g., each of the four steps above) is made repeatable based on curve fitting or a defined step - specific tolerance margin. A time reference can be defined to monitor the repeatability of different variables.

[0064] In some examples, a reference frame for process parameter monitoring is established for subsequent steps or cycles by a previous or initial step or cycle, as opposed to being established, for example, by set values and error bands in a conventional method. In fact, some examples are agnostic (i.e., ignore) with respect to conventional set values or error bands and operate based on repeating the previous cycle or step in a given substrate manufacturing process. For example, the parameters of an ALD cycle may be adjusted by trial and error to form a desired pattern on a substrate. Even if one does not necessarily know or be able to identify what the absolute values of those parameters as successful examples are, in the examples of the present disclosure, the processing chamber can be configured to simply repeat successful cycles based on data derived from curve fitting or a definition by a tolerance margin specific to the step.

[0065] FIG. 3 is a graph 300 depicting an exemplary curve fitting of 30 ALD cycles. The chamber pressure (y-axis) was monitored at time increments of 100 milliseconds (ms) during each ALD cycle. The elapsed time from the start to the end of each ALD cycle ranges from about 1 second or less, and as shown, chamber pressure data was collected for each time increment during that period. A set of curves 306 representing the chamber pressure vary between peaks 302 and valleys 304 of the data during successive steps (e.g., dose, purge, conversion, and purge) within each ALD cycle. In the example shown, each ALD cycle begins with a dose step.

[0066] The curve fitting operation is performed for each curve 306 in the set of curves, and as a result, consecutive or subsequent ALD cycles repeat the previous cycle. Each step within an ALD cycle can also repeat the previous cycle. In this regard, for one (or more) time increments associated with that step, a step-specific tolerance margin can be established. The tolerance margin can be based on variability data observed or set for the type shown in graph 400 of FIG. 4. Here, for each time increment, the variability (or tolerance margin) of the chamber pressure can be observed. A relatively wide (or loose) tolerance margin 402 can correspond to, for example, a relatively wide set of chamber pressures represented by the set of curves at 0.4 time increments in FIG. 3. Conversely, a relatively narrow (or tight) tolerance margin 404 can correspond to, for example, a relatively narrow (or equal) set of chamber pressures represented by the set of curves at 0.1 time increments in FIG. 3. For each time increment during a step or within an ALD cycle, the variability or tolerance margin can be defined. One or more curves 306 within the set of curves can also be defined, either alone or based on the tolerance margin. One or more sets of curves 306 and / or step-specific tolerance margins can be established and used, for example, for monitoring and repeatability of steps and cycles during an ALD cycle in a substrate processing operation.

[0067] FIG. 5 shows, for example, a flowchart of exemplary operations of a method 500 for monitoring steps or cycles in an ALD process. In some examples, operation 502 includes defining a time reference (start) for an ALD cycle (e.g., opening of a particular dose valve). In some examples, operation 502 can also include defining a "golden curve" or golden value for a series of data points (e.g., time increments). In some examples, operation 502 further includes defining the variability or tolerance margin for each data point (time increment).

[0068] In some examples, operation 504 includes collecting parameter data. In some examples, the parameter data is collected at regular intervals (e.g., at so-called trigger points) based on a certain collection frequency (e.g., 1 Hz, 10 Hz, 100 Hz, or 1000 Hz).

[0069] In some examples, operation 506 may include data comparison, for example, comparing data for various parameters such as chamber pressure, precursor manifold pressure, purge pressure, RF reflected power, and RF forward power. The parameter values for a particular time increment (trigger point) are compared to a golden curve or golden value. In some examples, the data comparison can be performed in real time, at the end of a given cycle or step, or at the end of a given substrate manufacturing process. This may be user-defined or based on process needs or optimization.

[0070] In some examples, operation 508 includes generating an algorithm that enables the repeatability of steps and cycles. Exemplary algorithms can include (or be based on) one or more of curve fitting, standard deviation from a golden value, or minimum or maximum ranges in variability or tolerance margins. Other algorithmic factors can also be used.

[0071] In some examples, operation 510 includes reporting (e.g., generating the output of curve fitting or generating a variability report). In some examples, operation 510 may include identifying and / or performing corrective actions, for example, to adjust process parameters based on a variability report or to fit process parameters within a set of fitted curves or tolerance margins. Some examples may include not taking action if the process parameters are within a set of fitted curves or tolerance margins. Accordingly, a tool or process chamber warning may or may not be generated.

[0072] Variations in process parameters can also make process control and monitoring in other areas difficult. For example, variations between substrates (or between lots) can occur due to the accumulation of chamber heat during substrate processing. Variations between tools can occur due to differences in pump efficiency. Conventionally, the main efforts to control variations have focused on monitoring the performance of individual devices. Examples of devices and their related parameters include MFC flow where the device error limit is set to 1% of the flow. The MFC flow is monitored to operate within its limit during substrate processing. The device and parameters can further include valve timing (e.g., ALD valve timing monitored to operate with an open time of 50 ms and a closed time of 70 ms). In another example, the valve may be set to switch between the open and closed positions in 25 ms. In other examples, a thermocouple may be used to monitor deviations within a set range to control pedestal temperature. Control of RF power can include monitoring forward power and reflected power. These devices typically have inherent performance or response limitations that can cause chamber control failures or make chamber conditions random or variable. Insufficient tightness of monitoring limits and the large number of process factors and device limitations can cause this undesirable effect.

[0073] Furthermore, referring to FIG. 6, unlike many PECVD processes where parameters such as gas flow, pressure, and RF power generally remain constant throughout the deposition process, some parameters in each cycle continuously change during the ALD process (cycle). Table 600 in FIG. 6 shows exemplary steps and related parameters in the ALD cycle. These steps can include, as illustrated, dose, post-dose purge, RF power application, and purge.

[0074] Some examples herein are aimed at solving such problems and include matching measured parameters as opposed to monitoring device parameters. Thus, some examples are configured to enable matching the performance of devices and / or chambers, e.g., for a given set of process steps or cycles and across a set of process chambers within a processing tool. For example, the pressure fluctuations or pressures generated during each ALD step can be matched cycle-by-cycle or across the tool and chambers by adjusting one or more process parameters that occur during the individual steps of the cycle. For example, a given pressure setpoint in an individual step (or cycle) of a substrate manufacturing process can be matched (or iterated) across the step or cycle by dynamic and real-time adjustment of the gas flow.

[0075] Similarly, level variations in the precursor manifold pressure can affect the deposition rate (depR) and, in some examples, can be an important variable in chamber matching. On the other hand, the precursor manifold pressure can be affected by one or more exemplary factors such as the precursor gas (e.g., Ar) push flow, the precursor flow (or ampoule temperature), the pump efficiency by shunting, and the chamber outlet or shunt timing. As described above, existing techniques attempt to control and monitor device setpoints such as MFC flow, valve opening and closing times, power to the RF generator, etc. to control chamber parameters to the setpoints. Variability between substrates, lots, and tools is controlled by using and monitoring device parameters. It is assumed that matching is possible by controlling the device parameters. However, the actual conditions within the processing chamber or gas line are not controlled or matched within the time limits imposed by the ALD steps and cycles.

[0076] Some examples include dynamic process control and monitoring for chamber matching. By way of example, achieving matching chamber performance can be accomplished by matching measurement parameters instead of device parameters. In some examples, a “golden curve” or golden value may represent, for example, a set of desired parameter values for performing a given or desired substrate formation within a chamber. The chamber may be operated using control variables set to “golden” conditions or “golden” values. In some examples, the “golden curve” may be based on parameter data derived from one step or cycle and may then be used as a reference or reference point for repeating that step or cycle or for matching parameter values or control variables across, for example, chambers, substrates, and tools.

[0077] Table 700 of FIG. 7 shows an exemplary set of golden values for relevant steps in an ALD cycle. The rows for pressure and gas flow each include two values, namely the “golden” value (simply denoted as pressure or flow) and the “modified” pressure or flow. In some examples, during operation, the pressure at each step is modified such that the monitored pressure curve matches the golden value of the golden chamber. Similarly, the gas flow at each step can be modified to match the golden flow value or golden pressure value. An adjustment in one parameter can cause a direct change in that parameter or an indirect change in another parameter related to the first parameter (e.g., a change in chamber gas flow affects the corresponding change in the pressure of the chamber gas flow).

[0078] In some examples, the pressure variations or pressures present in the first step of an ALD cycle are made to match in subsequent steps (or cycles) by adjusting a pressure setpoint, and thus, in some examples, the process is controlled to match this based on previous process parameters rather than by adjustment to a given device parameter. In some examples, the pressure matching adjustment is done for each individual step or cycle. In some examples, the pressure matching is done directly or indirectly, for example, by matching previous gas flows.

[0079] Figure 8 shows an exemplary matching operation in the integration method 800. In graph 802, the tool A pressure curve 804 in the first step or cycle does not match the tool B pressure curve 806 in the second step or cycle, and it can be seen that the curves of their respective values do not match and are offset relative to each other. In the illustrated example, it can be said that the tool B pressure value of curve 806 is delayed relative to the tool A pressure value of curve 804. In some examples, as shown in graph 808, an adjustment is made to chamber parameters such as gas flow or pressure parameters so that the two curves 804 and 806 substantially match. Steps or cycles that match exactly are useful in ensuring the repeatability of chamber conditions and consistency in substrate manufacturing, and as a result, the accuracy in the production of substrate formations and semiconductor devices is improved.

[0080] In other examples, the precursor manifold pressure curves of two different chambers during an ALD cycle may not match even though all relevant devices on the tool are functioning within specifications. Variability between chambers can result from differences in pump efficiency, differences in precursor manifold pressure and valve timing of chamber valves, gas line temperature, etc. Figure 9 depicts an example of the difference between chambers. For example, in graph 900, the difference in pressure for the chambers can be seen during period 902. In graph 904, the difference in pressure for each (between chambers) is observed. The illustrated pressure curves 906 and 908 have the magnitude of time and pressure offset (not matching) relative to each other.

[0081] In some examples, the issue of chamber-to-chamber matching is addressed. In this regard, Table 1000 of FIG. 10 shows a set of parameters for adjusting the precursor manifold pressure in an ALD cycle. Exemplary parameters include precursor gas flow (e.g., argon Ar), precursor flow, ampule inlet / outlet valve timing, and chamber inlet / outlet valve timing. Exemplary parameters may each include a corresponding golden value or golden curve. Other parameters may also be used.

[0082] The effect or influence of the control on the precursor manifold pressure can be imparted by making specific adjustments in some parameters, as shown, for example, in the comment section of Table 1000. For example, the argon push flow can affect the magnitude of the precursor manifold pressure and move the pressure curve vertically. In other words, when matching the precursor manifold pressure curve established in the second or subsequent step (or cycle) with the corresponding curve established (or generated) in the previous step or cycle, if it appears "lower" than the previous curve in the matching graph (e.g., graph 802), the pressure curves in subsequent cycles (such as the third and fourth cycles) can be corrected to move upward by appropriately adjusting the argon precursor push flow. Therefore, the conditions of the processing chamber can be configured to be essentially constant by monitoring the parameter values obtained from the previous step or cycle, rather than based on fixed set values or error bands.

[0083] FIG. 11 shows an exemplary operation in the integration method 1100. Here, changes in parameters such as the push flow (per ALD cycle), ampule temperature, and valve timing are utilized to establish matching precursor manifold pressures 1102 and 1104 in each step or cycle of the ALD process. Furthermore, the matched precursor manifold pressures can match the deposition rates between the dependent chambers.

[0084] Exemplary operations in a method of dynamic process control can include selecting a control variable to match (e.g., pressure), determining a parameter to change to modify the control variable, determining a control increment of the parameter, and calculating a deviation from the golden curve of the control variable. Referring to FIG. 12, examples of control operations in a method 1200 for dynamic process control within a chamber include, in operation 1202, identifying a control variable (e.g., pressure) and establishing a golden curve or value for it; in operation 1204, identifying a first adjustment parameter that can affect the value of the identified control variable; in operation 1206, determining the magnitude or step increment in the identified parameter for the control variable; in operation 1208, selecting a parameter increment that has the greatest potential impact on the control variable; in operation 1210, determining whether the control variable is within specifications; if "yes", in operation 1212, repeating operations 1206 and 1208; if "no", in operation 1212, applying the selected parameter increment; in operation 1214, determining whether the control variable is within specifications after application of the selected parameter increment, and so on. If "yes", method 1200 includes repeating operations 1206 and 1208. If "no", method example 1200 includes, in operation 1216, identifying a second adjustment parameter (e.g., gas flow setting) that can affect the value of the identified control variable, and in operations 1218 - 1226, repeating the control operations summarized above for the second adjustment parameter. When the selected increments of the first and second adjustment parameters establish the golden value of the identified control variable, in 1218, the chamber is optimized for pressure matching.

[0085] Some examples include methods for determining deviations from a golden curve in dynamic process control. Examples of operations in this method include defining a time reference at the start of an ALD cycle (e.g., a digital output signal for opening a dosing valve of a chamber). Thereafter, all data points collected at regular intervals (e.g., 1 ms, 10 ms, 50 ms, 100 ms) are compared with the corresponding data in the golden curve. The deviation or error limit from the golden curve can be predefined based on experiments or user specifications. When a parameter is optimized for a particular measurement, e.g., chamber pressure, the effect of changes in the parameter on other control variables, e.g., precursor manifold pressure, is checked and adjusted as needed. Table 1300 in FIG. 13 shows examples of parameters for making corrections to enable the alignment of different control variables.

[0086] Some examples in this specification achieve alignment between substrates, lots, and chambers by matching dynamic parameter control to actual measured values, e.g., chamber pressure, gas line pressure, temperature, supplied RF power. Some examples include the optimization of parameters to minimize measurement deviations from a golden curve or golden value. The optimized parameters may vary from chamber to chamber. The optimized parameters may vary in each accumulation to account for other factors (e.g., drift). Some examples include software features (described in more detail below) for performing parameter optimization for process control. This process control method can be executed regularly or intermittently, e.g., during continuous operation, or at tool startup or scheduled maintenance. Parameter optimization can be executed with an assigned priority and, in some examples, based on the control variable that has the most impact on a given process.

[0087] In a further aspect, some examples include self - learning techniques for advanced monitoring of processes, particularly ALD and CVD processes. These techniques are aimed at addressing monitoring problems that can occur with conventional methods for controlling conventional equipment. For example, MFCs are typically monitored near setpoint values within a set error range. In valve timing control, typically, the opening and closing times of ALD valves are monitored, and the ALD valves can switch between open and closed states during a given cycle or process in a short period of 50 ms. RF generation is typically measured and controlled based on forward and reflected power and then monitored within an error band. Chamber and other temperatures are controlled and monitored near setpoint values within an error band / percentage.

[0088] (In contrast to devices) Conventional efforts to monitor process parameters include monitoring chamber pressure near setpoint values within a set error band. The error band is often set large enough to ignore the inherent variations that occur during an ALD cycle. As such, these efforts are not capable of faithfully monitoring more detailed aspects or chamber conditions during an ALD cycle. There is an increasing case for using detailed and deep chamber control to enable the fabrication of high aspect ratio nano - sized substrate formations and semiconductor devices. Further, the precursor manifold and burst purge pressure are also typically monitored around the band. Conventional error bands are typically set too wide to capture smaller variations during an ALD cycle and thus pose a problem in the same way as described above.

[0089] Regarding the supplied RF power, a one-time check of the RF power simply checks whether the high frequency is on or off after an RF strike. A voltage-current (VI) sensor monitors the RF power during the plasma "on" step and monitors it simply at a frequency higher than the RF power (e.g., at 1 kHz). Thus, in a broad sense, existing methods are based on limited or error band settings and typically perform passive or "silent" monitoring methods based on limited data. Generally, the monitoring band is very wide and cannot even address, let alone cope with, the strict process control issues constantly required in today's semiconductor manufacturing. As yet another drawback, the same monitoring band is applied to all tools and no tool-to-tool modification or customization is performed. Typically, customization is ad hoc and performed manually. Typically, in the prior art, little or no performance or comparison between substrates or tools is performed regardless of whether accumulation is considered, after preventive maintenance, or after a hardware change.

[0090] As described above, the ALD process can be regarded as a multi-step process. Referring to FIG. 14, a typical ALD cycle 1400 includes four main steps: dose 1402, purge 1404, conversion 1406, and purge 1408. In some of these embodiments, each step of the ALD cycle and subsequent cycles of a given ALD process are monitored individually for different variables. The monitored variables are matched in each ALD step and / or cycle using curve fitting or defined error margins in a "smart" self-learning monitoring process. All ALD cycles are made repeatable and the repeatability of parameters such as chamber pressure, precursor manifold pressure, temperature, etc. can be monitored (and made repeatable) in all cycles or steps.

[0091] FIG. 15 shows operations 1502 to 1522 in a self - learning monitoring method 1500 as an example. Method 155 includes, in operation 1502, defining one or more parameters to be monitored, such as valve timing, VI sensor, RF forward direction, RF reflection direction, pressure (such as chamber pressure, precursor manifold pressure, or burst pressure), and other parameters. Initial data is compared with start values and collected to define a basic goodness, and may also include a golden value or a golden curve. The golden curve may be, for example, as shown in FIG. 3. Operation 1504 includes data analysis and logging of aspects such as mean value, standard deviation, out - of - control rate (OOC), and other statistical values. Operation 1504 may further include updating performance data with new data such as the derived statistical values. Operation 1506 may include monitoring chamber performance based on the deviation from one or more statistical values such as mean value performance. Operation 1508 may include determining whether the deviation is within the device limits. If "yes", previous operations such as operation 1504 may be repeated. Operation 1510 may include determining whether the deviation is outside the device limits. If "no", operation 1512 includes issuing a user review warning. As a result of the review, if "no problem", previous operations within the method may be repeated as illustrated. If "problem", operation 1522 includes issuing a tool warning. Operation 1514 may include saving tracker data at regular intervals (definable by the user). Operation 1516 includes comparing performance data with tracker data to identify data drift. Operation 1518 includes identifying whether there is data drift based on the comparison. If "yes", operation 1520 includes issuing a report to obtain user input and corrective actions.

[0092] An example of a self - learning monitoring method may include collecting data from multiple systems and defining in - chamber performance regarding performance alignment between chambers and between tools. FIG. 16 shows exemplary operations 1602 - 1612 in a data collection method 1600. In operation 1602, multiple systems (e.g., processing chambers) are identified. In operation 1604, the monitoring system obtains the actual performance or the golden curve value of a parameter from a selected module within the identified processing chamber (or tool). An acquisition frequency is specified. The acquisition frequency may be, for example, a frequency defined by the user, daily, based on tracker data, or weekly.

[0093] In operation 1606, statistical process control (SPC) measurements are obtained from each chamber or tool. Operation 1608 includes performing a comparison and analysis of between - chamber (or between - tool) performance for each of the relevant parameters and defining an average value performance and a standard deviation for each chamber (e.g., chambers 1 - 3 in the figure). Operation 1608 may include generating a warning regarding outlier performance based on user - defined criteria (e.g., 3σ) and establishing a correlation with the acquired SPC data. Operation 1610 includes performing an overall between - chamber (or between - tool) comparison and analysis, and in operation 1612, issuing a tool warning if appropriate.

[0094] In some examples, the smart self-learning monitoring system includes a machine learning component that creates a monitoring process based on a machine learning model. The component may include a component that performs preprocessing of data. The preprocessing component receives, for example, training data supplied from a processing chamber, or a set of golden values or golden curves (such as of the type shown in FIG. 3). The preprocessing component performs preprocessing of the training data, including, for example, applying a MapReduce function or a similar function to the training data. Next, a feature extraction component extracts a plurality of features (such as process parameters) from the preprocessed training data and supplies these features to a machine learning algorithm. The extracted features may be related to one or more of the control variables described above. In some examples, the machine learning algorithm learns weights assigned to each of the features and applies these weights to a function. This function and the learned weights may be included in or constitute the machine learning model described above. One or more machine learning models are stored in a file system and are read out as needed to perform analysis of chamber performance or process monitoring.

[0095] The machine learning algorithm can be selected from a variety of numerous candidates of supervised or unsupervised machine learning algorithms. Examples of supervised machine learning algorithms include artificial neural networks, Bayesian networks, instance-based learning, support vector machines, random forests, linear classifiers, quadratic classifiers, k-nearest neighbor methods, decision trees, and hidden Markov models. Examples of unsupervised machine learning algorithms include the EM (expectation maximization) algorithm, vector quantization, and information bottleneck methods. In one example of an embodiment, a binary logistic regression model is used. Binary logistic regression deals with situations where the observed results for a dependent variable can have only two types as candidates. Logistic regression is used to predict the odds that one or the other case is true based on the values of independent variables (predictors). In a further example of an embodiment, a boosted tree gradient descent process is used for machine learning.

[0096] The functions included in the machine learning model are evaluable at runtime to generate a process alignment score. The alignment score is a prediction about the likelihood that attempts to match conditions in multiple systems, based on evaluating various parameters and applying the weights of features learned by a machine learning algorithm to those features, will result in alignment success. In some examples, the predicted alignment may include an output that includes a hybrid result or a parameter adjustment that results in alignment between chambers or between tools with improved reliability.

[0097] In further examples, systems and methods are provided that include hardware and software for determining gas line fill times. In some exemplary deposition or etching systems, gas is supplied from a gas box and includes several valves and filters between the gas source and the deposition or etching chamber. In this regard, typically, an MFC, MFC inlet and outlet valves, filters, and chamber inlet valves are used. The travel time of gas from a particular valve to the chamber can be defined as the gas line fill time.

[0098] In some examples, the gas line fill time can be affected by certain factors. These factors can include the MFC response and ramp time. In some examples, the ramp time takes 1 - 3 seconds for the corresponding gas flow to reach within + / - 2% of the set value. Factors for the fill time can also include the valve opening time. In some examples, the valve opening time is in the range of several milliseconds including air pressure delay and other delays. Some examples include valve opening times in the range of <100 ms for pneumatically actuated valves. Other factors can include the conductance of the gas line including valve conductance, or the pressure drop across a filter. Basically, the gas velocity depends on the differential pressure, and the differential pressure is affected or determined by the conductance of the gas line. For deposition and etching processes that can provide sufficient time for a given gas line fill, any delay in gas line fill will have no significant impact on the deposition or etching process, if any.

[0099] As further described above, the ALD process can include multiple steps that result in film formation on a substrate surface. These steps can include a dose step where precursor gas molecules adhere to the substrate surface, a post-dose purge to remove excess precursor gas from the chamber, an application of RF power to generate plasma and convert the monolayer of gas molecules adsorbed on the surface into a film, and an RF purge to remove reaction by-products. The gases used in these steps are typically emitted from various manifolds.

[0100] Currently, the gas line fill time is neither measured nor monitored. The opening and closing times of the valve can be monitored using advance frequency (kHz) monitoring. The ALD valve can include an optical sensor that detects the position of the valve throttle. The opening and closing timing of the valve is monitored by the time difference between a command (digital input, DI) to the pneumatic bank that operates the valve and the feedback of the optical sensor (digital output, DO) that senses the movement of the throttle.

[0101] For example, referring to the valve arrangement 1700 shown in FIG. 17, in a deposition or etching system, gas is typically supplied from a gas box, and several valves and filters can be provided between one or more gas sources and the deposition / etching chamber. Typically, the valves include MFCs, MFC inlet and outlet valves, filters, and chamber inlet valves. The transit time of gas from a particular valve to the chamber can be defined as the gas line fill time. The gas line fill time depends on the MFC response and ramp time. In some examples, the valves typically take 1 - 3 seconds for the gas flow to reach within + / - 2% of the set value. The valve opening time typically occurs in the range of a few milliseconds. Including the compressed dry air delay and other delays, the valve opening time can be less than 100 ms for pneumatically actuated valves. The valve conductance can include the pressure drop across the filter. Basically, the gas velocity depends on the differential pressure, while the differential pressure is controlled by the conductance of the gas line. In deposition and etching processes where sufficient time can be ensured for filling the gas line, such gas line fill delays may not pose a problem in the deposition / etching process, but on the other hand, there may be cases where the gas line fill delay is unacceptable. Considering the need to shorten the ALD cycle time for high throughput (and deep substrate forming ability), the MFC ramp time or delay in the substrate processing operation is unacceptable. Therefore, in some examples, the gas line fill time is determined to exclude potential causes of delay or instability. Based on the fill time thus determined, the MFC that supplies the relevant gas is configured to operate in a continuous or consistent manner.

[0102] The presence of a split (when gas is not required, the chamber is bypassed around the gas and made to flow, for example, into an exhaust line or a so-called fore line) can also affect the gas line fill time. FIG. 18 is a schematic diagram of a shared split configuration 1800 for a gas manifold. The gas is split away from the chamber and typically results in a relatively long gas line fill. FIG. 19 is a schematic diagram of an individual station split configuration 1900 for a gas manifold. The gas is split towards the chamber and typically results in a relatively short gas line fill.

[0103] In an ALD process, the individual step time can include the gas line fill time from the closest outlet valve to the process chamber. Thus, in some exemplary embodiments, it shows a response to the important but complex requirement of providing a measurable gas line fill time that is consistent between tools.

[0104] Some exemplary embodiments automatically measure the gas line fill time at various times (e.g., at startup, after preventive maintenance, or at regular intervals). The current measurement value is automatically compared to previous measurement values, and any deviation in the gas line fill time is reported. More generally, the measurement value is also compared to gas line fill times derived from other tools, allowing determination of tool-to-tool variation and tool conformance or status.

[0105] In some examples, a chamber vacuum gauge is used to measure or monitor the gas line fill time. For example, the time difference between the opening of the manifold outlet valve and the increase in chamber pressure is used in some examples to measure or calculate the gas line fill time. Some exemplary methods include establishing a base pressure or a constant pressure within the process chamber supplied by a gas pump. This method includes closing the pump throttle and / or the divert valve to isolate the chamber from the pump. The gas flow is set to divert from the appropriate manifold or gas line for which the line fill time is being measured. This method further includes opening the gas manifold outlet valve and closing the divert valve to measure the increase in chamber pressure. Typically, some initial delay occurs before the chamber pressure begins to rise. The pressure ramp curve used in some examples can be used to calculate the gas line fill time or the delay. In other words, the initial delay of the pressure ramp indicates the gas line fill time to the chamber. The calculated delay (fill time) can be factored into control and monitoring systems and methods, including gas inflow algorithms, to improve the steady-state operation and chamber matching operation of the substrate processing chamber.

[0106] Here, with reference to graph 2000 of FIG. 20, an example method for determining the gas line fill time will be described. Graph 2000 includes a line 2002 representing the movement (closing or opening) of a gas valve that supplies gas to or removes gas from the process chamber. In this case, as an example of the valve, it is a divert valve as shown in the graph key, and the divert valve controls the gas in the line that bypasses the chamber. When the divert valve is closed, the gas does not divert from the chamber and instead is directed to enter the chamber. Gas enters the chamber and the pressure inside the chamber rises. Typically, the closing of the divert valve is performed simultaneously with the opening of the chamber supply valve to enable supply to the chamber. In any case, in this example, the gas line fill time of the divert valve is established.

[0107] Line 2002 represents the physical closure of the bleed valve from the open (bleed) position at position "1" on the y-axis to the closed position at "0" on the same axis. As described above, when the bleed valve closes, the pressure in the chamber rises, and this rise is represented by the pressure line 2004 in graph 2000. After a certain period, the slope of line 2004 becomes constant, i.e., it represents a uniform or steady rise in gas pressure occurring in response to a constant or steady inflow of gas into the chamber. The slope of line 2004 increases as the gas flow increases. The gas line fill time is determined by extrapolating the slope of line 2004 until it intersects the x-axis. This intersection is shown at position 2006 in graph 2000, and the period represented by the time line 2008 extending between the intersection 2006 and the final closure of the bleed valve shown at position 2010 is the gas line fill time. In other words, there is a ramp-up or delay time (i.e., the gas line fill time) between the closure of the bleed valve (i.e., the complete inflow or maximum fill of gas into the chamber) and the point at which the pressure in the chamber starts to rise at a constant rate.

[0108] In some embodiments, the gas decay time is determined. The closure of the gas outlet valve does not mean an immediate stop of the gas flow into the chamber; there is a delay both when the valve closure is commanded and when the valve physically reaches full closure, and there are also delays caused by other components. Even when the valve is closed, the gas downstream of the valve that is already in the gas line still continues into the chamber.

[0109] In some examples, the method of establishing the gas decay (or residence) time may include one or more of the following operations. First, the chamber is implemented at a base pressure or a constant pressure. The chamber is supplied by a throttle and a slit valve that control the inflow of gas from a supply pump. The supply throttle and the slit valve supplying the chamber are closed to isolate the chamber from the supply pump. The gas flow is set to divert from the manifold or gas line for which the gas decay time is to be measured. The gas manifold outlet valve is opened and the divert valve is closed. The increase in chamber pressure is measured. Thereafter, the divert valve is opened and the gas outlet valve is closed. This method further includes measuring the time it takes for the chamber pressure to maximize or stabilize. The time it takes for the chamber pressure to maximize or stabilize is a measure of the time (i.e., the gas decay or residence time) required for the inflow of gas into the chamber to stop after a valve closure instruction is sent to the outlet valve.

[0110] FIG. 21 shows a graph 2100 representing an exemplary gas line fill time and an exemplary gas decay time. The line 2102 in the graph represents the opening and closing of the valve that supplies the processing chamber. The valve is closed during the period indicated at 2104 and opened during the period 2106 to allow gas to flow into the chamber and increase the pressure inside the chamber. The valve is closed again at 2108. The chamber pressure is measured by a chamber vacuum gauge and is represented by the pressure line 2110. As described above, the gas line fill time can be determined from the period indicated at 2112 or Δ (i.e., the period between the intersection of the extrapolation of the pressure gradient and the x-axis and the moment of the valve opening instruction at 2102). The gas decay time is determined from the time it takes to stabilize after the valve is closed at 2108. This gas decay time is more clearly visualized in the enlarged view at 2114 (80 ms in this example).

[0111] Accordingly, some examples provide a method for measuring gas line fill time using chamber pressure, and a method for measuring gas decay (or residence) time using chamber pressure. These values may be integrated into automated parameter measurement, chamber control, and dynamic monitoring processes and software for alignment techniques. Fill and decay (residence) values can be important in monitoring tool-to-tool (or chamber-to-chamber) variability and daily drift in gas line filling or decay over time. Such methods are applicable to product lines that employ multiple step procedures (or cycles) for deposition or etching and require very fast step and cycle times. At faster step and cycle times, gas line fill and decay times can become important. The present embodiment enables those measurements and monitoring.

[0112] Exemplary embodiments may include a method. Referring to FIG. 22, a method 2200 for monitoring a processing cycle in an ALD semiconductor manufacturing process includes, in operation 2202, defining a time reference for an ALD cycle based on repetitive operations in the manufacturing process; in operation 2204, accessing a golden curve including a series of parameter values for a series of data points at periodic time increments based on the time reference; in operation 2206, accessing variability or tolerance margins for each data point in the golden curve; in operation 2208, collecting parameter data based on a periodic time increment of one cycle in the ALD manufacturing process; in operation 2210, dynamically monitoring whether a parameter value of the parameter data is within the variability or tolerance margin at a certain data point; and in operation 2212, adjusting the manufacturing process such that parameter values in subsequent cycles match associated parameter values in the golden curve based on a determination that the parameter value has deviated from the variability or tolerance margin.

[0113] In some examples, the repetitive operations that form the basis of the time reference include opening and closing a particular valve that supplies a processing chamber.

[0114] In some examples, the parameter data is collected at regular intervals based on a collection frequency, and the collection frequency is within the range of 0 to 1 Hz, 1 to 10 Hz, 10 to 100 Hz, or 100 to 1000 Hz.

[0115] In some examples, the regular intervals are based on trigger points in the ALD manufacturing process, and each trigger point defines or is based on a time point in a step within the ALD manufacturing process.

[0116] In some examples, the operation further includes comparing the parameter data collected at the trigger point with a corresponding parameter data set in the golden curve.

[0117] In some examples, the parameter data includes parameter values related to one or more of a precursor manifold pressure, a purge pressure, a conversion manifold pressure, a chamber pressure, a gas flow, an RF reflected power, and an RF forward power.

[0118] Referring to FIG. 23, a method 2300 for monitoring a processing cycle in a semiconductor manufacturing process includes, in operation 2302, identifying parameters of the manufacturing process, in operation 2304, generating a first curve of parameter values including a first parameter value based on a first cycle of the manufacturing process, in operation 2306, identifying a second value of a parameter for a second cycle of the manufacturing process based on the curve generated for the first cycle, and in operation 2308, adjusting the manufacturing process to match the first parameter value with the second parameter value.

[0119] In some examples, the operation further includes generating a second curve including a plurality of parameter values derived from a second cycle of the manufacturing process, and curve fitting the second curve of parameter values with the first curve of parameter values.

[0120] In some examples, the curve fitting operation includes applying a series of parameter value curves based on or derived from the third and subsequent cycles of the manufacturing process to the first or second curve to generate a golden curve that defines a set of golden parameter values for the manufacturing process.

[0121] In some examples, each cycle of the manufacturing process includes a plurality of steps of an ALD process, and the operation further includes matching the parameter values at each step of the ALD process with the parameter values in the set of golden parameter values.

[0122] In some examples, the golden curve includes the golden parameter values for each step of the ALD process.

[0123] In some examples, the identified parameters are associated with control variables of the manufacturing process, and the operation further includes identifying a matching value between the value of the control variable in the second cycle and the value of the control variable in the first cycle using the first and second parameter values directly or indirectly, and adjusting the manufacturing process to match the first parameter value with the second parameter value.

[0124] Referring to FIG. 24, a machine learning method 2400 for monitoring processing steps in a semiconductor manufacturing cycle includes, in operation 2402, generating a set of reference golden parameter values for each step in a series of iterative cycles based on parameter data collected from a processing chamber, in operation 2404, generating a machine learning model based on the set of reference golden values, and in operation 2406, using the machine learning model to match the parameter values in the second cycle in a series of iterative cycles with the corresponding parameter values in the first cycle in the series of iterative cycles.

[0125] In some examples, the semiconductor manufacturing process is an ALD process, and the steps of each cycle in a series of iterative cycles include continuous steps including a dose step, a purge step, a conversion step, and a purge step.

[0126] In some examples, the operation further includes generating a golden curve including parameter value data at each continuous step and using the golden curve as training data for a machine learning model.

[0127] In some examples, the operation further includes repeating and matching each cycle in a series of iterative cycles by matching the parameter values at each continuous step of a second cycle with the corresponding parameter values of each continuous step of a first cycle based on a machine learning model.

[0128] In some examples, the parameter includes one or more of a precursor manifold pressure, a purge pressure, a chamber pressure, a gas flow, a chamber temperature, an RF reflected power, and an RF forward power.

[0129] In some examples, the operation further includes collecting performance data from a processing chamber, identifying a drift of the performance data, generating performance drift data, and incorporating the drift data into training data for a machine learning model.

[0130] Referring to FIG. 25, a method 2500 for monitoring a processing cycle in a semiconductor manufacturing process includes, in operation 2502, identifying parameters of the manufacturing process and, in operation 2504, generating a first curve of parameter values including a first parameter value based on a first cycle of the manufacturing process, where the parameter values include a gas line fill time of a line supplying a processing chamber.

[0131] In some examples, the operation further includes calculating a gas line fill time based on a pressure ramp-up time between opening a valve in a line supplying the process chamber and establishing a subsequent constant pressure increase.

[0132] In some examples, the operation further includes identifying a second value of a parameter for a second cycle of a manufacturing process based on a curve generated for a first cycle, generating a second curve including a plurality of parameter values derived from the second cycle of the manufacturing process, and curve fitting the second curve of parameter values to the first curve of parameter values.

[0133] In some examples, the curve fitting operation includes fitting a series of parameter value curves based on or derived from third and subsequent cycles of the manufacturing process to the first or second curve to generate a golden curve defining a set of golden parameter values for the manufacturing process.

[0134] In some examples, each cycle of the manufacturing process includes a plurality of steps in an ALD process, and the operation further includes matching parameter values at each step of the ALD process to parameter values in the set of golden parameter values.

[0135] In some examples, the golden curve includes golden parameter values for each step of the ALD process.

[0136] In some examples, the identified parameter is associated with a control variable of the manufacturing process, and the operation further includes using the first and second parameter values directly or indirectly to identify a matching value between the value of the control variable in the second cycle and the value of the control variable in the first cycle, and adjusting the manufacturing process to match the first parameter value to the second parameter value.

[0137] FIG. 26 is a block diagram illustrating an example of a machine 2600 capable of controlling one or more exemplary embodiments processes described herein. In alternative embodiments, machine 2600 may operate as a stand-alone device or may be connected (e.g., network-connected) to other machines. In a network deployment, machine 2600 may operate in the capacity of a server machine, a client machine, or both, in a server-client network environment. In one example, machine 2600 may operate as a peer machine in a peer-to-peer (P2P) (or other distributed) network environment. Further, although only a single machine 2600 is shown, the term "machine" shall also be construed to include any collection of machines that individually or jointly execute a set (or multiple sets) of instructions to perform any one or more of the methodologies discussed herein via cloud computing, software as a service (SaaS), or other computer cluster configurations. In some examples, as shown in FIG. 26, a non-transitory machine-readable medium includes instructions 2624 that, when read by machine 2600, cause the machine to be controlled to operate in a manner that includes at least the non-limiting exemplary operations summarized above and described herein.

[0138] The examples described in this specification may include or be operable by logic, some components, or mechanisms. A circuit set is a collection of circuits implemented in a tangible entity that includes hardware (e.g., simple circuits, gates, logic, etc.). The membership of a circuit set can flexibly accommodate the passage of time and the variability of the underlying hardware. A circuit set includes members that can perform specific operations during operation, either alone or in combination. In one example, the hardware of a circuit set may be fixedly designed (e.g., hardwired) to perform a specific operation. In one example, the hardware of a circuit set includes physically modifiable (e.g., by magnetic, electrical, movable placement of invariant mass particles, etc.) computer-readable media for encoding instructions for a specific operation, and may include variably connected physical components (e.g., execution units, transistors, simple circuits, etc.). When connecting physical components, the underlying electrical characteristics of the hardware components are changed (e.g., from insulator to conductor, or vice versa). The instructions enable the embedded hardware (e.g., execution unit or loading mechanism) to create members of the circuit set within the hardware via variable connections and execute a part of a specific operation during operation. Thus, the computer-readable media are communicatively coupled to other components of the circuit set when the device is operating. In one example, any of the physical components may be used by multiple members of multiple circuit sets. For example, during operation, an execution unit may be used by a first circuit of a first circuit set at one point in time and reused by a second circuit within the first circuit set or by a third circuit within a second circuit set at another point in time.

[0139] A machine (e.g., a computer system) 2600 can include a hardware processor 2602 (e.g., a central processing unit (CPU), a hardware processor core, or any combination thereof), a graphics processing unit (GPU) 2632, a main memory 2604, and a static memory 2606, and some or all of which can communicate with each other via an interlink (e.g., a bus) 2608. The machine 2600 can further include a display device 2610, an alphanumeric input device 2612 (e.g., a keyboard), and a user interface (UI) navigation device 2614 (e.g., a mouse). In one example, the display device 2610, the alphanumeric input device 2612, and the UI navigation device 2614 can be a touch screen display. The machine 2600 can further include a mass storage device (e.g., a drive unit) 2616, a signal generation device 2618 (e.g., a speaker), a network interface device 2620, and one or more sensors 2630 (such as a global positioning system (GPS) sensor, a compass, an accelerometer, or another sensor). The machine 2600 can include an output controller 2628 such as a serial (e.g., universal serial bus (USB)), parallel, or other wired or wireless (e.g., infrared (IR), near field communication (NFC), etc.) connection to communicate with or control one or more peripheral devices (e.g., a printer, a card reader, etc.).

[0140] The large-capacity storage device 2616 may include a machine-readable medium 2622. Stored in this machine-readable medium 2622 are one or more sets of data structures or instructions 2624 (e.g., software) that embody one or more of the technologies or functions described herein or are utilized by one or more of the technologies or functions described herein. Also as shown, the instructions 2624 may be present, in whole or at least in part, within the main memory 2604, within the static memory 2606, within the hardware processor 2602, or within the GPU 2632 during execution by the machine 2600. In one example, any one of the hardware processor 2602, the GPU 2632, the main memory 2604, the static memory 2606, or the large-capacity storage device 2616, or any combination thereof, may constitute the machine-readable medium 2622.

[0141] The machine-readable medium 2622 is shown as a single medium, but the term "machine-readable medium" may include a single medium configured to store one or more instructions 2624, or multiple media (e.g., centralized or distributed databases, and / or associated caches and servers).

[0142] The term "machine-readable medium" can include any medium that can store, encode, or carry instructions 2624 for execution by a machine 2600 and cause the machine 2600 to perform any one or more of the techniques of this disclosure, or any medium that can store, encode, or carry a data structure used by such instructions 2624 or a data structure related to such instructions 2624. Non-limiting examples of machine-readable media can include solid state memories, optical media, and magnetic media. In one example, a mass machine-readable medium includes a machine-readable medium 2622 having a plurality of particles with invariant (e.g., stationary) mass. Thus, a mass machine-readable medium is not a signal propagating temporarily. Specific examples of mass machine-readable media can include non-volatile memories such as semiconductor memory devices (e.g., electrically programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM)) and flash memory devices, magnetic disks such as internal hard disks and removable disks, magneto-optical disks, and CD-ROM and DVD-ROM disks. Further, the instructions 2624 can be transmitted or received using a transmission medium, through a communication network 2626 via a network interface device 2620.

[0143] Examples have been described with reference to specific exemplary embodiments or methods, but it will be apparent that various modifications and changes can be made to these embodiments without departing from the broader scope of the embodiments. Accordingly, the specification and drawings are to be considered in an illustrative rather than a limiting sense. The accompanying drawings, which form a part of this specification, illustrate, by way of example and not limitation, specific embodiments in which the subject matter may be practiced. The illustrated embodiments are described in sufficient detail to enable those skilled in the art to practice the teachings disclosed herein. Other embodiments may be utilized and other embodiments may be derived from the teachings disclosed herein without departing from the scope of the present disclosure. Accordingly, this detailed description is not to be construed in a limiting sense, and the scope of the various embodiments is defined only by the appended claims and all ranges of equivalents to which such claims are entitled.

[0144] Such embodiments of the subject matter of the present invention may be referred to herein individually and / or collectively by the term "invention", which is merely a matter of convenience and is not intended to voluntarily limit the scope of this application to any single invention or inventive concept (if more than one are actually disclosed). Accordingly, while specific embodiments have been illustrated and described herein, it should be recognized that any configuration calculated to achieve the same purpose may be an alternative to the specific embodiments shown. The present disclosure is intended to cover any and all adaptations or variations of the various embodiments. Combinations of the above-described embodiments and other embodiments not specifically described herein will be apparent to those skilled in the art upon consideration of the above description. Also, the present disclosure can be realized in the following forms. [Form 1] A system for monitoring a processing cycle in a manufacturing process, a processing chamber for the manufacturing process, and one or more controllers configured to perform a process monitoring operation, wherein the operation identifies parameters of the manufacturing process, and includes generating a first curve of parameter values including a first parameter value based on a first cycle of the manufacturing process, the parameter values including a gas line filling time of a line supplying the processing chamber. System. [Form 2] The system according to Form 1, wherein the operation further includes calculating the gas line filling time based on a pressure ramp-up time between an opening of a valve in the line supplying the processing chamber and an establishment of a subsequent constant pressure increase. [Form 3] The system according to Form 2, wherein the operation based on the curve generated for the first cycle, identifies a second value of the parameter for a second cycle of the manufacturing process, generates a second curve including a plurality of parameter values derived from the second cycle of the manufacturing process, and curve-fits the second curve of parameter values to the first curve of parameter values. System. [Form 4] The system according to Form 3, wherein the curve-fitting operation includes generating a golden curve that defines a set of golden parameter values for the manufacturing process by fitting a curve of a series of parameter values based on or derived from the third and subsequent cycles of the manufacturing process to the first or second curve. System. [Form 5] The system according to Form 4, wherein each cycle of the manufacturing process includes a plurality of steps in an ALD process, and the operation further includes matching parameter values in each step of the ALD process with parameter values in the set of golden parameter values. System. [Form 6] The system according to Form 4, wherein the golden curve includes golden parameter values for each step of the ALD process. [Form 7] The system according to Form 5, wherein the identified parameter is associated with a control variable of the manufacturing process, The operation is, using the first and second parameter values directly or indirectly to identify a matching value between the value of the control variable in the second cycle and the value of the control variable in the first cycle, and adjusting the manufacturing process to match the first parameter value with the second parameter value further comprising a system. [Form 8] A method for monitoring a processing cycle in a manufacturing process, comprising: identifying parameters of the manufacturing process, generating a first curve of parameter values including a first parameter value based on a first cycle of the manufacturing process, the parameter values including a gas line filling time of a line supplying the processing chamber. Method. [Form 9] The method according to Form 8, wherein the operation is, calculating the gas line filling time based on a pressure ramp-up time between an opening of a valve in the line supplying the processing chamber and an establishment of a subsequent constant pressure increase. further comprising a method. [Form 10] The method according to Form 9, wherein the operation is, based on the curve generated for the first cycle, identifying a second value of the parameter for a second cycle of the manufacturing process, generating a second curve including a plurality of parameter values derived from the second cycle of the manufacturing process, and curve fitting the second curve of parameter values to the first curve of parameter values. further comprising a method. [Form 11] The method according to Form 10, wherein the curve fitting operation is, generating a golden curve that defines a set of golden parameter values for the manufacturing process by fitting a curve of a series of parameter values based on or derived from the third and subsequent cycles of the manufacturing process to the first or second curve. [Form 12] The method according to Form 11, wherein each cycle of the manufacturing process includes a plurality of steps in an ALD process, The operation is, matching parameter values in each step of the ALD process to the parameter values in the set of golden parameter values A method further comprising this. [Form 13] The method according to Form 11, wherein the golden curve includes golden parameter values for each step of the ALD process. [Form 14] The method according to Form 12, wherein the identified parameter is associated with a control variable of the manufacturing process, The operation is, using the first and second parameter values directly or indirectly to identify a matching value between the value of the control variable in the second cycle and the value of the control variable in the first cycle, and adjusting the manufacturing process to match the first parameter value to the second parameter value A method further comprising this. [Form 15] A machine-readable medium which, when read by a machine, causes the machine to, identify parameters of a manufacturing process, execute an operation of generating a first curve of parameter values including a first parameter value based on a first cycle of the manufacturing process, the parameter values including a gas line filling time of a line supplying the processing chamber, A machine-readable medium. [Form 16] The medium according to Form 15, wherein the operation is, calculating the gas line filling time based on a pressure ramp-up time between opening of a valve in the line supplying the processing chamber and subsequent establishment of a constant pressure increase A medium further comprising this. [Form 17] The medium according to Form 16, wherein the operation is, identifying a second value of the parameter for a second cycle of the manufacturing process based on the curve generated for the first cycle, generating a second curve including a plurality of parameter values derived from the second cycle of the manufacturing process, and curve fitting the second curve of parameter values to the first curve of parameter values A medium further comprising this. [Form 18] The medium according to Form 17, wherein the curve fitting operation is, generating a golden curve that defines a set of golden parameter values for the manufacturing process by fitting a curve of a series of parameter values based on or derived from the third and subsequent cycles of the manufacturing process to the first or second curve. [Form 19] The medium according to form 18, wherein each cycle of the manufacturing process includes a plurality of steps in an ALD process, The operation is, matching the parameter values in each step of the ALD process with the parameter values in the set of golden parameter values further comprising a medium. [Form 20] The medium according to form 18, wherein the golden curve includes golden parameter values for each step of the ALD process. [Form 21] The medium according to form 15, wherein the identified parameter is associated with a control variable of the manufacturing process, The operation is, using the first and second parameter values directly or indirectly to identify a matching value between the value of the control variable in the second cycle and the value of the control variable in the first cycle, and adjusting the manufacturing process to match the first parameter value with the second parameter value further comprising a medium.

Claims

1. A system for monitoring a processing cycle in a manufacturing process, a processing chamber for the manufacturing process, and one or more controllers configured to perform a process monitoring operation, wherein the operation identifies parameters of the manufacturing process, includes generating a first curve of parameter values including a first parameter value based on a first cycle of the manufacturing process, wherein the parameter values include a gas line filling time of a line supplying the processing chamber, the operation further includes calculating the gas line filling time based on a pressure ramp-up time between an opening of a valve in the line supplying the processing chamber and an establishment of a subsequent constant pressure increase, wherein the manufacturing process is an atomic layer deposition (ALD) process, system.

2. The system according to claim 1, wherein the operation identifies a second value of the parameter for a second cycle of the manufacturing process based on the curve generated for the first cycle, generates a second curve including a plurality of parameter values derived from the second cycle of the manufacturing process, and curve fits the second curve of parameter values to the first curve of parameter values further includes, system.

3. The system according to claim 2, wherein the curve fitting operation fits a curve of a series of parameter values based on or derived from the third and subsequent cycles of the manufacturing process to the first or second curve to generate a golden curve defining a set of golden parameter values for the manufacturing process includes, system.

4. The system according to claim 3, wherein each cycle of the manufacturing process includes a plurality of steps in an ALD process, the operation wherein the operation matches parameter values in each step of the ALD process with parameter values in the set of golden parameter values further includes, system.

5. The system according to claim 3, wherein the golden curve includes golden parameter values for each step of the ALD process, system.

6. The system according to claim 4, wherein the identified parameters are associated with control variables of the manufacturing process, the operation Identifying a matching value between the value of the control variable in the second cycle and the value of the control variable in the first cycle by directly or indirectly using the first and second parameter values, and Adjusting the manufacturing process to match the first parameter value with the second parameter value A system further comprising.

7. A method for monitoring a processing cycle in a manufacturing process, comprising Identifying parameters of the manufacturing process, Generating a first curve of parameter values including a first parameter value based on a first cycle of the manufacturing process, wherein the parameter value includes a gas line filling time of a line supplying a processing chamber, The method further comprises Calculating the gas line filling time based on a pressure ramp-up time between opening of a valve in the line supplying the processing chamber and establishment of a subsequent constant pressure increase Including further, The manufacturing process is an atomic layer deposition (ALD) process,[[]] Method.

8. The method according to claim 7, wherein the method further comprises Based on the curve generated for the first cycle, Identifying a second value of the parameter for a second cycle of the manufacturing process, Generating a second curve including a plurality of parameter values derived from the second cycle of the manufacturing process, and Curve fitting the second curve of parameter values to the first curve of parameter values A method further comprising.

9. The method according to claim 8, wherein the curve fitting operation comprises Fitting a curve of a series of parameter values based on or derived from the third and subsequent cycles of the manufacturing process to the first or second curve to generate a golden curve defining a set of golden parameter values for the manufacturing process.

10. The method according to claim 9, wherein each cycle of the manufacturing process includes a plurality of steps in an ALD process, The operation further comprises Matching the parameter values in each step of the ALD process with the parameter values in the set of golden parameter values A method further comprising.

11. The method according to claim 9, wherein the golden curve includes golden parameter values for each step of the ALD process.

12. The method according to claim 10, wherein the identified parameter is associated with a control variable of the manufacturing process, the operation is, using the first and second parameter values directly or indirectly to identify a matching value between the value of the control variable in the second cycle and the value of the control variable in the first cycle, and adjusting the manufacturing process to match the first parameter value with the second parameter value further comprising a method.

13. A machine-readable medium that, when read by a machine, causes the machine to monitor a processing cycle in a manufacturing process, and when read by the machine, causes the machine to, identify parameters of the manufacturing process, execute an operation of generating a first curve of parameter values including a first parameter value based on a first cycle of the manufacturing process, the parameter values including a gas line filling time of a line supplying a processing chamber for the manufacturing process, the operation further comprising calculating the gas line filling time based on a pressure ramp-up time between an opening of a valve in the line supplying the processing chamber and an establishment of a subsequent constant pressure increase, the manufacturing process is an atomic layer deposition (ALD) process, machine-readable medium.

14. The medium according to claim 13, wherein the operation is, identifying a second value of the parameter for a second cycle of the manufacturing process based on the curve generated for the first cycle, generating a second curve including a plurality of parameter values derived from the second cycle of the manufacturing process, and curve fitting the second curve of parameter values to the first curve of parameter values further comprising a medium.

15. The medium according to claim 14, wherein the curve fitting operation is, including generating a golden curve that defines a set of golden parameter values for the manufacturing process by fitting a curve of a series of parameter values based on or derived from the third and subsequent cycles of the manufacturing process to the first or second curve.

16. The medium according to claim 15, wherein each cycle of the manufacturing process includes a plurality of steps in an ALD process, the operation is, Matching the parameter values in each step of the ALD process with the parameter values in the set of golden parameter values A medium further comprising this.

17. The medium according to claim 15, wherein the golden curve includes golden parameter values for each step of the ALD process.

18. The medium according to claim 13, wherein the identified parameter is associated with a control variable of the manufacturing process, The operation is Using the first and second parameter values directly or indirectly to identify a matching value between the value of the control variable in the second cycle of the manufacturing process and the value of the control variable in the first cycle, and Adjusting the manufacturing process so that the first parameter value matches the second parameter value A medium further comprising this.

19. A system for monitoring a processing cycle in a manufacturing process, A processing chamber for the manufacturing process, and Including one or more controllers configured to perform a process monitoring operation, The operation is Identifying the parameters of the manufacturing process, Generating a first curve of parameter values including a first parameter value based on a first cycle of the manufacturing process, the parameter values including the gas line filling time of the line supplying the processing chamber, The gas line filling time includes a valve opening time, The manufacturing process is an atomic layer deposition (ALD) process, A system.

20. A method for monitoring a processing cycle in a manufacturing process, Identifying the parameters of the manufacturing process, Generating a first curve of parameter values including a first parameter value based on a first cycle of the manufacturing process, the parameter values including the gas line filling time of the line supplying the processing chamber, The gas line filling time includes a valve opening time, The manufacturing process is an atomic layer deposition (ALD) process.

21. A machine-readable medium that, when read by a machine, causes the machine to monitor a processing cycle in a manufacturing process, and when read by the machine, causes the machine to Identify the parameters of the manufacturing process, A command for causing an operation of generating a first curve of parameter values including a first parameter value based on a first cycle of the manufacturing process is included, where the parameter values include a gas line filling time of a line supplying a processing chamber for the manufacturing process. The gas line filling time includes a valve opening time. The manufacturing process is an atomic layer deposition (ALD) process, a machine-readable medium.

Images